Appendix A: Summary of evidence from surveillance

2018 surveillance of Epilepsies: diagnosis and management (2012) NICE guideline CG137

Summary of evidence from surveillance

1.1 Principle of decision making

Recommendations in this section of the guideline 1.1.1 Healthcare professionals should adopt a consulting style that enables the child, young person or adult with epilepsy, and their family and/or carers as appropriate, to participate as partners in all decisions about their healthcare, and take fully into account their race, culture and any specific needs. [2004]

Initial intelligence gathering identified NICE Principle of decision making guideline CG138 Patient experience in adult NHS services which covers the components of 2018 surveillance summary a good patient experience. No relevant evidence was identified. Impact statement 2014 surveillance summary NICE guideline CG138 provides a set of No relevant evidence was identified. recommendations on making sure that adults using NHS services have the best possible

experience of care, as such NICE guideline Topic expert feedback and additional CG137 should cross-refer to the guideline; and information consideration should be given to withdrawing the current recommendation. No topic expert feedback was relevant to this evidence. New evidence identified that may change current recommendations.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 1 of 81 1.2 Coping with epilepsy

Recommendations in this section of the guideline 1.2.1 Children, young people and adults with epilepsy and their families and/or carers should be empowered to manage their condition as well as possible. [2004] 1.2.2 Adults should receive appropriate information and education about all aspects of epilepsy. This may be best achieved and maintained through structured self-management plans. [2004] 1.2.3 In children and young people, self-management of epilepsy may be best achieved through active child-centred training models and interventions. [2004] 1.2.4 Healthcare professionals should highlight the Expert Patients Programme* to children, young people and adults with epilepsy who wish to manage their condition more effectively. [2004, amended 2012]

* This web address has changed since the recommendation was published in 2004 and has been updated.

chronic epilepsy (n = 83). This reported that Coping with epilepsy intervention participants significantly improved, relative to controls on the Epilepsy 2018 surveillance summary Self-Management Scale, Epilepsy Self-Efficacy No relevant evidence was identified. Scale and Quality of Life in Epilepsy-31. [1]

2014 surveillance summary Impact statement No relevant evidence was identified. The evidence supports current recommendation content to offer structured Topic expert feedback self-management plans. A topic expert highlighted 1 RCT which compared a self-management intervention New evidence is unlikely to change guideline with treatment-as-usual with adults with recommendations.

1.3 Information

Recommendations in this section of the guideline 1.3.1 Children, young people and adults with epilepsy and their families and/or carers should be given, and have access to sources of, information about (where appropriate):  epilepsy in general  diagnosis and treatment options  medication and side effects  seizure type(s), triggers and seizure control  management and self-care  risk management

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 2 of 81  first aid, safety and injury prevention at home and at school or work  psychological issues  social security benefits and social services  insurance issues  education and healthcare at school  employment and independent living for adults  importance of disclosing epilepsy at work, if relevant (if further information or clarification is needed, voluntary organisations should be contacted)  road safety and driving  prognosis  sudden death in epilepsy (SUDEP)  status epilepticus  lifestyle, leisure and social issues (including recreational drugs, alcohol, sexual activity and sleep deprivation)  family planning and pregnancy  voluntary organisations, such as support groups and charitable organisations, and how to contact them. [2004] 1.3.2 The time at which this information should be given will depend on the certainty of the diagnosis, and the need for confirmatory investigations. [2004] 1.3.3 Information should be provided in formats, languages and ways that are suited to the child, young person or adult's requirements. Consideration should be given to developmental age, gender, culture and stage of life of the person. [2004] 1.3.4 If children, young people and adults, and their families and/or carers, have not already found high-quality information from voluntary organisations and other sources, healthcare professionals should inform them of different sources (using the Internet, if appropriate: see, for example, the website of the Joint Epilepsy Council of the UK and Ireland). [2004] 1.3.5 Adequate time should be set aside in the consultation to provide information, which should be revisited on subsequent consultations. [2004] 1.3.6 Checklists should be used to remind children, young people and adults, and healthcare professionals, about information that should be discussed during consultations. [2004] 1.3.7 Everyone providing care or treatment for children, young people and adults with epilepsy should be able to provide essential information. [2004] 1.3.8 The child, young person or adult with epilepsy and their family and/or carers as appropriate should know how to contact a named individual when information is needed. This named individual should be a member of the healthcare team and be responsible for ensuring that the information needs of the child, young person or adult and/or their family and/or carers are met. [2004] 1.3.9 The possibility of having seizures should be discussed, and information on epilepsy should be provided before seizures occur, for children, young people and adults at high risk of developing seizures (such as after severe brain injury), with a learning disability, or who have a strong family history of epilepsy. [2004] 1.3.10 Children, young people and adults with epilepsy should be given appropriate information before they make important decisions (for example, regarding pregnancy or employment). [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 3 of 81 Sudden unexpected death in epilepsy (SUDEP) 1.3.11 Information on SUDEP should be included in literature on epilepsy to show why preventing seizures is important. Tailored information on the person's relative risk of SUDEP should be part of the counselling checklist for children, young people and adults with epilepsy and their families and/or carers. [2004] 1.3.12 The risk of SUDEP can be minimised by:  optimising seizure control  being aware of the potential consequences of nocturnal seizures. [2004] 1.3.13 Tailored information and discussion between the child, young person or adult with epilepsy, their family and/or carers (as appropriate) and healthcare professionals should take account of the small but definite risk of SUDEP. [2004] 1.3.14 Where families and/or carers have been affected by SUDEP, healthcare professionals should contact families and/or carers to offer their condolences, invite them to discuss the death, and offer referral to bereavement counselling and a SUDEP support group. [2004]

with refractory epilepsy are at lower risk of Information SUDEP if they are treated with effective doses of adjunctive AEDs. [3] 2018 surveillance summary A Cochrane review evaluating the Topic expert feedback effectiveness of interventions in preventing Topic experts highlighted the Practice Sudden Unexpected Death in Epilepsy guideline summary: Sudden unexpected death (SUDEP) in people with epilepsy, only found 1 in epilepsy incidence rates and risk factors: case-control study at serious risk of bias Report of the Guideline Development, (n=154 cases of SUDEP and 616 controls). The Dissemination, and Implementation study showed that nocturnal supervision, a Subcommittee of the American Academy of supervising person sharing the same bedroom, Neurology and the American Epilepsy Society. or taking special precautions such as using a listening device have a protective effect A topic expert said that parents frequently ask against SUDEP, which is independent of about overnight monitoring of their child’s seizure control. [2] seizures. There are now many seizure detection devices available, some of which 2014 surveillance summary (Emfit monitor, epilepsy smartwatches) have undergone clinical testing (no references A systematic review and meta-analysis of 112 provided). However, there is no guidance on RCTs (21,224 participants) investigating what to advise parents regarding monitoring whether receiving effective doses of anti- their child overnight. epileptic drugs (AEDs) reduced the risk of sudden unexpected death in epilepsy (SUDEP) Impact statement in patients with refractory epilepsy found that compared with patients allocated to placebo, This section of the guideline that relates to patients assigned to effective doses of AEDs SUDEP should be updated. The Cochrane had a lower incidence of definite and probable review reported that there was 'very low- SUDEP and a reduced frequency of all causes quality evidence of a preventative effect for of death. The evidence suggests that adults nocturnal supervision against SUDEP', the

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 4 of 81 current recommendation notes the importance with epilepsy that seizure freedom, particularly of being aware of the 'potential consequences freedom from generalized tonic-clonic seizures, of nocturnal seizures' but does not mention is strongly associated with decreased risk of nocturnal supervision, while the evidence is SUDEP. very low quality, this may have an impact on the recommendation. The practice guideline is New evidence identified that may change in line with current recommendations, but also current recommendations. highlights additional details such as recommending that clinicians inform people

1.4 Following a first seizure

Recommendations in this section of the guideline 1.4.1 Children, young people and adults presenting to an Accident and Emergency department following a suspected seizure should be screened initially. This should be done by an adult or paediatric physician with onward referral to a specialist* when an epileptic seizure is suspected or there is diagnostic doubt. [2004] 1.4.2 Protocols should be in place that ensure proper assessment in the emergency setting for children, young people and adults presenting with an epileptic seizure (suspected or confirmed). [2004] 1.4.3 The information that should be obtained from the adult and/or family or carer after a suspected seizure is contained in appendix D. [2004] 1.4.4 The information that should be obtained from the child or young person and/or parent or carer after a suspected seizure is contained in appendix D. [2004] 1.4.5 It is recommended that all adults having a first seizure should be seen as soon as possible** by a specialist in the management of the epilepsies to ensure precise and early diagnosis and initiation of therapy as appropriate to their needs. [2004] 1.4.6 It is recommended that all children and young people who have had a first non-febrile seizure should be seen as soon as possible** by a specialist in the management of the epilepsies to ensure precise and early diagnosis and initiation of therapy as appropriate to their needs. [2004] 1.4.7 At the initial assessment for a recent onset seizure, the specialist should have access to appropriate investigations. [2004] 1.4.8 In a child, young person or adult presenting with an attack, a physical examination should be carried out. This should address their cardiac, neurological and mental status, and should include a developmental assessment where appropriate. [2004] 1.4.9 Essential information on how to recognise a seizure, first aid, and the importance of reporting further attacks should be provided to a child, young person or adult who has experienced a possible first seizure, and their family/carer/parent as appropriate. This information should be provided while the child, young person or adult is awaiting a diagnosis and should also be provided to their family and/or carers. [2004]

* For adults, a specialist is defined throughout as a medical practitioner with training and expertise in epilepsy. For children and young people, a specialist is defined throughout as a paediatrician with training and expertise in epilepsy.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 5 of 81 ** The Guideline Development Group considered that with a recent onset suspected seizure, referrals should be urgent, meaning that patients should be seen within 2 weeks.

Following a first seizure Impact statement No new information was identified at any 2018 surveillance summary surveillance review. The absence of new No relevant evidence was identified. evidence indicates that there is no need to update this section of the guideline. 2014 surveillance summary No relevant evidence was identified. No new evidence. Recommendations are unlikely to change. Topic expert feedback No relevant evidence was identified.

1.5 Diagnosis

Recommendations in this section of the guideline 1.5.1 The diagnosis of epilepsy in adults should be established by a specialist medical practitioner with training and expertise in epilepsy. [2004] 1.5.2 The diagnosis of epilepsy in children and young people should be established by a specialist paediatrician with training and expertise in epilepsy. [2004] 1.5.3 Children, young people and adults and their families and/or carers should be given an opportunity to discuss the diagnosis with an appropriate healthcare professional. [2004] 1.5.4 A detailed history should be taken from the child, young person or adult and an eyewitness to the attack, where possible, to determine whether or not an epileptic seizure is likely to have occurred. [2004] 1.5.5 The clinical decision as to whether an epileptic seizure has occurred should then be based on the combination of the description of the attack and different symptoms. Diagnosis should not be based on the presence or absence of single features. [2004] 1.5.6 It may not be possible to make a definite diagnosis of epilepsy. If the diagnosis cannot be clearly established, further investigations (see section 1.6) and/or referral to a tertiary epilepsy specialist* (see recommendation 1.10.2) should be considered. Follow-up should always be arranged. [2004] 1.5.7 Where non-epileptic attack disorder is suspected, suitable referral should be made to psychological or psychiatric services for further investigation and treatment. [2004] 1.5.8 Prospective recording of events, including video recording and written descriptions, can be very helpful in reaching a diagnosis. [2004]

* In this recommendation, 'centre' has been replaced with 'specialist' for consistency across recommendations.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 6 of 81 A retrospective study of 595 patients with Diagnosis either ES, PNES, physiologic nonepileptic seizure-like events (PSLE), mixed PNES plus ES 2018 surveillance summary or inconclusive monitoring assessed frequency of “positive complaints” on ROS questionnaires Clinical decision rules as an early screening tool for distinguishing A systematic review examined use of clinical between seizure groups. Binomial regression decision rules (CDRs) in epilepsy. Five studies showed some patient groups (PNES, mixed considered to be moderate quality were PNES and ES) had a significantly greater included; meta-analysis of two validated CDRs number of symptoms compared with PSLE or found high initial accuracy levels with isolated ES. Multivariate logistic regression, sensitivity of approximately 80% of above, controlling for sex and the number of medical which "tended to diminish significantly" in comorbidities demonstrated that the validation studies. Pooled estimates of prospective accuracy to differentiate PNES sensitivity and specificity showed wide 95% from ES was not significantly higher than confidence and prediction intervals which may assuming that all patients had ES (76% versus restrict their value in clinical practice. [4] 70%, p>0.1). The authors concluded that the consistency of ROS responses was ‘neither Questionnaires/psychometric tests accurate nor specific enough to be used solely A systematic review of 9 studies focused on as an early screening tool for PNES’. [7] use of screening questionnaires to diagnose A retrospective study of 44 patients with epilepsy in adult subjects compared with either PNES or ES assessed following clinical evaluation. The authors noted that completion of a standardised ROS to assess studies had a high risk of bias. In individuals the value of multiple complaints as a marker of with a lifetime history of epilepsy ranges of PNES. Multivariate analysis of covariance 81.5-100% for sensitivity and 65.6-99.2% for showed a significantly greater number of specificity were reported. In individuals with complaints in patients with PNES compared active epilepsy sensitivity and specificity with ES in the ROS questionnaire, whereby a ranged from 48.6-100% and 73.9-99.9%, threshold level of 17% of positive complaints respectively. [5] resulted in sensitivity and specificity values of A retrospective study with patients with either 85% and 78% respectively for differentiating psychogenic non-epileptic seizures (PNES) between PNES and ES. [8] (n=30) or epilepsy (n=30) matched on age and A study used the Minnesota Multiphasic sex, assessed the value of a review of system Personality Inventory-2-Restructured Form (ROS) questionnaire in distinguishing between (MMPI-2-RF) to differentiate between seizure seizure groups. The questionnaire consists of types in 169 individuals with either epileptic 10 general yes/no questions about the seizures (ES) or PNES. In PNES, the odds of presence or absence of any abnormality in scale elevations for certain behaviours body systems. The mean (+/- SD) ROS (negative mood, suicidal ideation and somatic response for the presence of any abnormality concerns) were 2 to 5 times more likely. was 2.43 (+/-1.33) for the PNES group and Gender differences were found with female 1.50 (+/-0.94) for the epilepsy group (p=0.01). PNES patients having 3-6 times greater odds A cut-off point of three positive ROS was able of such scale elevations, while male PNES to differentiate the two conditions (p=0.01; patients had odds of 5-15 times more likely. OR: 6, 95% confidence interval: 1.48-24.29)". Positive and negative predictive values varied The authors concluded that "multiple between 53.66-84.62% and 47.52-90.91% complaints in the ROS questionnaire argues in respectively and false discovery and omission favour of PNES compared with epilepsy". [6]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 7 of 81 rates ranged from 15.38- 50% and 9.09- differential diagnosis of PNES and ES. A 52.47% respectively for scales with increased multivariate logistic regression model odds of elevation in PNES subjects. [9] differentiated PNES only from ES only with a prospective accuracy of 78% and area under In a cross-sectional study, the diagnostic value the curve of 79%. Medication and number of of Multiphasic Personality Inventory co-morbidities were greater predictors of Minnesota 2 (MMPI-2) was assessed in 209 seizure type compared with a particular patients with either PNES, PNES combined comorbidity. However, certain co-morbidities with seizures, or patients with seizures only; were significantly associated with PNES the reference was video-EEG. In a clinical (asthma, chronic pain, migraines p<0.01) and scales model (comprising sex, hypochondriasis ES (diabetes and non-metastatic neoplasm, and paranoia), sensitivity and specificity were p<0.01). [12] 77.1% and 76.8% respectively, with a percentage of correct classification of 76.8%. A total of 201 children with history of one or In a content scales model comprising sex, HEA more episodes of loss of consciousness and (health concerns) and FRS (fears), sensitivity diagnosed as having neurally mediated syncope and specificity were 65.7% and 78.0%, or epilepsy were enrolled in a prospective respectively, with a percentage of correct study to determine the use of Calgary and classification of 75.9%. No significant modified Calgary score in differential diagnosis. differences were found between models. The There were significant differences in median authors concluded that "MMPI-2 had Calgary score between syncope [-4.00 (-6, 1)] moderate validity for the diagnosis of PNES in and epilepsy [2 (-3, 5)] (z = -11.63, P < 0.01) patients referred to an epilepsy unit", however and in median modified Calgary score between utilisation of content scales did not syncope [-4.00 (-6, 1)] and epilepsy [3 (-3, 6)] (z significantly enhance clinical scales results. [10] = -11.71, P < 0.01). Scores greater than or A cohort study assessed a self-reported equal to 1 produced high sensitivity and symptom questionnaire in 386 patients with specificity for both Calgary score (91.46 and diagnosed epilepsy, 308 patients with 95.80% respectively) and modified Calgary diagnosed PNES, and 371 patients with scores (92.68 and 96.64% respectively), both diagnosed syncope. The analysis concentrated indicating a diagnosis of epilepsy. The on transient loss of consciousness (TLOC). sensitivity and specificity of modified Calgary PNES patients had a significantly higher score and Calgary score did not show number and frequency of TLOC related significant differences (P > 0.05). [13] symptoms compared with epilepsy or syncope patients (p <0.001) and pairwise logistic 2014 surveillance summary regression analysis correctly classified 91% of No relevant evidence was identified. patients with epilepsy versus those with syncope, 94% of those with PNES versus those Topic expert feedback with syncope, and 77% of those with epilepsy Topic experts advised that the versus those with PNES. Comparable findings recommendation needed updating due to new were found following multinomial logistic International League Against Epilepsy (ILAE) regression analysis. [11] guidelines on diagnosis and classification of epilepsy and ILAE Official Report: A practical Other differential diagnostic studies clinical definition of epilepsy. These reports A study in 1365 patients with ES only, possible consider that epilepsy may be diagnosed after PNES only, mixed PNES and ES, PSLE or single seizure if there is a high risk of further inconclusive monitoring evaluated comorbid seizures occurring; additional discussion may diagnoses and medication history for

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 8 of 81 be required as to when an individual may be of system and personality questionnaires, considered not to have epilepsy. clinical decision rules and investigations of co- morbidity in diagnosing epilepsy and One topic expert also suggested that there is differentiating between different types of an inadequate focus on non-epileptic seizures, seizures. and noted these patients can make up to one- third of patients referred to secondary care The changes in agreed ILAE criteria for with refractory epilepsy. However, non- diagnosing epilepsy should be considered in epileptic seizures are outside the scope for this the update of the guideline. guideline. New evidence identified that may change Impact statement current recommendations. There is mixed evidence concerning the sensitivity and specificity of screening, review

1.6 Investigations

Recommendations in this section of the guideline 1.6.1 Information should be provided to children, young people and adults and families and/or carers as appropriate on the reasons for tests, their results and meaning, the requirements of specific investigations, and the logistics of obtaining them. [2004] 1.6.2 All investigations for children should be performed in a child-centred environment. [2004] Electroencephalogram (EEG) 1.6.3 Children, young people and adults requiring an EEG should have the test performed soon after it has been requested*. [2004] 1.6.4 An EEG should be performed only to support a diagnosis of epilepsy in adults in whom the clinical history suggests that the seizure is likely to be epileptic in origin. [2004] 1.6.5 An EEG should be performed only to support a diagnosis of epilepsy in children and young people. If an EEG is considered necessary, it should be performed after the second epileptic seizure but may, in certain circumstances, as evaluated by the specialist, be considered after a first epileptic seizure. [2004] 1.6.6 An EEG should not be performed in the case of probable syncope because of the possibility of a false-positive result. [2004] 1.6.7 The EEG should not be used to exclude a diagnosis of epilepsy in a child, young person or adult in whom the clinical presentation supports a diagnosis of a non-epileptic event. [2004] 1.6.8 The EEG should not be used in isolation to make a diagnosis of epilepsy. [2004] 1.6.9 An EEG may be used to help determine seizure type and epilepsy syndrome in children, young people and adults in whom epilepsy is suspected. This enables them to be given the correct prognosis. [2004] 1.6.10 In children, young people and adults presenting with a first unprovoked seizure, unequivocal epileptiform activity shown on EEG can be used to assess the risk of seizure recurrence. [2004] 1.6.11 For children, young people and adults in whom epilepsy is suspected, but who present diagnostic difficulties, specialist investigations should be available. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 9 of 81 1.6.12 Repeated standard EEGs may be helpful when the diagnosis of the epilepsy or the syndrome is unclear. However, if the diagnosis has been established, repeat EEGs are not likely to be helpful. [2004] 1.6.13 Repeated standard EEGs should not be used in preference to sleep or sleep-deprived EEGs. [2004] 1.6.14 When a standard EEG has not contributed to diagnosis or classification, a sleep EEG should be performed. [2004] 1.6.15 In children and young people, a sleep EEG is best achieved through sleep deprivation or the use of melatonin**. [2004, amended 2012] 1.6.16 Long-term video or ambulatory EEG may be used in the assessment of children, young people and adults who present diagnostic difficulties after clinical assessment and standard EEG. [2004] 1.6.17 Provocation by suggestion may be used in the evaluation of non-epileptic attack disorder. However, it has a limited role and may lead to false-positive results in some people. [2004] 1.6.18 Photic stimulation and hyperventilation should remain part of standard EEG assessment. The child, young person or adult and family and/or carer should be made aware that such activation procedures may induce a seizure and they have a right to refuse. [2004] Neuroimaging 1.6.19 Neuroimaging should be used to identify structural abnormalities that cause certain epilepsies. [2004] 1.6.20 MRI should be the imaging investigation of choice in children, young people and adults with epilepsy. [2004] 1.6.21 MRI is particularly important in those:  who develop epilepsy before the age of 2 years or in adulthood  who have any suggestion of a focal onset on history, examination or EEG (unless clear evidence of benign focal epilepsy)  in whom seizures continue in spite of first-line medication. [2004] 1.6.22 Children, young people and adults requiring MRI should have the test performed soon*. [2004] 1.6.23 Neuroimaging should not be routinely requested when a diagnosis of idiopathic generalised epilepsy has been made. [2004] 1.6.24 CT should be used to identify underlying gross pathology if MRI is not available or is contraindicated, and for children or young people in whom a general anaesthetic or sedation would be required for MRI but not CT. [2004] 1.6.25 In an acute situation, CT may be used to determine whether a seizure has been caused by an acute neurological lesion or illness. [2004] Other tests 1.6.26 Measurement of serum prolactin is not recommended for the diagnosis of epilepsy. [2004] 1.6.27 In adults, appropriate blood tests (for example, plasma electrolytes, glucose, calcium) to identify potential causes and/or to identify any significant co-morbidity should be considered. [2004] 1.6.28 In children and young people, other investigations, including blood and urine biochemistry, should be undertaken at the discretion of the specialist to exclude other diagnoses, and to determine an underlying cause of the epilepsy. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 10 of 81 1.6.29 A 12-lead ECG should be performed in adults with suspected epilepsy. [2004] 1.6.30 In children and young people, a 12-lead ECG should be considered in cases of diagnostic uncertainty. [2004] 1.6.31 In cases of diagnostic uncertainty, a referral to a cardiologist should be considered. [2004] Neuropsychological assessment 1.6.32 Neuropsychological assessment should be considered in children, young people and adults in whom it is important to evaluate learning disabilities and cognitive dysfunction, particularly in regard to language and memory. [2004] 1.6.33 Referral for a neuropsychological assessment is indicated:  when a child, young person or adult with epilepsy is having educational or occupational difficulties  when an MRI has identified abnormalities in cognitively important brain regions  when a child, young person or adult complains of memory or other cognitive deficits and/or cognitive decline. [2004]

* The Guideline Development Group considered that 'soon' meant being seen within 4 weeks. ** The licence for use of melatonin in the UK has changed since the recommendation was published in 2004. The recommendation has been updated accordingly and the footnote that contained the old information has been deleted.

children. Eleven children with epilepsy and 7 2018 surveillance summary control children were studied. Functional connectivity networks were significantly Electroencephalogram (EEG) different between children with epilepsy and controls. The system was able to diagnose Duration of EEG paediatric epilepsy with accuracy of 89%, A retrospective study assessed the effect of sensitivity of 82% and specificity of 100%. [15] limiting duration of continuous video EEG in an epilepsy monitoring unit on the diagnostic Specialist interpretation of EEG yield. Data from 2 epilepsy monitoring units A study assessed the intra-rater and inter-rater were analysed (n=560), one of which is open reliability of EEG interpretation. During two only from Monday to Friday, limiting the distinct time intervals, 6 clinical duration of observation to 5 days. The event neurophysiologists classified EEGs into one or capture rates were 74% for the institution more of 7 diagnostic categories and assigned a without time limits and 72% for the institution subjective confidence to their interpretations. with the 5-day time limit (statistical analysis of Each EEG was read by 3 people, each reader the outcome was not reported in the abstract). interpreted 150 unique studies, and 50 studies The time until first event was significantly were re-interpreted to generate intra-rater longer in the institution with the 5-day time data. The contribution of the EEGs, readers, limit than in the institution without a time limit; and the interaction between the EEGs and however, this was less than 5 days at both readers to interpretation variance were institutions; however the abstract did not analysed. Five of the six readers had a median report the p-value for this outcome. [14] confidence of 99% or higher. Intra-rater A study assessed a time varying approach for agreement ranged from 0.33 to 0.73 with an constructing functional connectivity networks aggregated value of 0.59. Inter-rater from EEG data for diagnosis of epilepsy in agreement ranged from 0.29 to 0.62, with an

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 11 of 81 aggregate value of 0.44. Variance due to the review of continuous EEG in patients admitted EEGs was 65.3%, due to readers was 3.9%, and to hospital (total 2,092 hours of EEG due to the interaction between readers and recording). Three neurophysiologists reviewed EEGs was 30.8%. [16] the reported findings of the first 30 minutes of 118 continuous EEGs, then used CSA to guide Automatic and computer-aided EEG subsequent review. Reviewers viewed 120 interpretation seconds of raw EEG data surrounding A study assessed a supervised machine suspicious CSA segments. The same learning approach to classify seizure and non- neurophysiologists performed independent seizure records using an open dataset (n=342). page-by-page conventional visual The machine learning approach improved on interpretation all continuous EEGs. existing studies by up to 10%, with sensitivity Independent conventional review by 2 of 93%, specificity of 94%, and an area under additional, more experienced the curve of 98%. However, the abstract did neurophysiologists was the gold standard. not describe the ‘existing studies’ used as the Reviewing 24 hours of EEG data was comparator. [17] significantly faster for compressed spectral array EEG data than for conventional EEG data. A study assessed quantitative EEG Longer review was necessary for both spectrogram (Persyst 12 EEG software) in compressed spectral array EEG data and distinguishing epileptic from non-epileptic conventional EEG data when seizures were events in people with paroxysmal non-epileptic detected, but the compressed review remained events (n=17), captured during EEG monitoring faster than conventional review. Compressed compared with a control group of patients with spectral array-guided review was sensitive for epileptic seizures of similar semiology (n=13). seizures (87.3%), periodic epileptiform Assessment of raw EEG was the gold standard discharges (100%), rhythmic delta activity against which epileptic and non-epileptic (97.1%), focal slowing (98.7%), generalised events were validated. Quantitative EEG slowing (100%), and epileptiform discharges spectrograms were interpreted as indicating a (88.5%). [19] seizure if it showed a visually significant change from baseline at the time of the Sleep or sleep-deprived EEGs clinically identified event. Eighty-two clinically A retrospective study assessed the value of identified paroxysmal events were analysed EEG containing sleep compared with an EEG of (46 non-epileptic and 36 epileptic). The awake time only in children aged 6 months to quantitative EEG spectrogram correctly 16 years (n=425 recordings). Recordings with classified 71% of non-epileptic events and 81% less than 10 minutes of awake time or at least of epileptic seizures. Additionally, rhythmicity 5 minutes of sleep were excluded. Additional spectrogram correctly classified 61% of non- yield of sleep was considered if at least one of epileptic events and 75%% epileptic events. the following was observed: appearance of Fast Fourier transform spectrogram correctly interictal epileptiform activity or increase by classified 65% of non-epileptic events and 69% more than 50%; or interictal epileptiform epileptic events. Asymmetry relative activity change in localization or morphology, spectrogram correctly classified 48% of non- seizure occurrence. Overall, 194 recordings epileptic events and 69% epileptic events; and (45.6%) showed an additional yield during integrated-amplitude EEG spectrogram sleep, which was significantly higher when correctly classified 59% of non-epileptic events interictal epileptiform activity was detected and 75% epileptic events. [18] during wakefulness. The yield was significantly A study assessed the use of compressed lower in recordings performed for non- spectral array EEG data in assisting expert epileptic referrals. [20]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 12 of 81 A retrospective cohort study examined sleep in 33% patients with a negative standard whether ambulatory EEG (aEEG) has similar EEG. [23] diagnostic accuracy as sleep deprived EEG A retrospective study used video-EEG analysis (sdEEG) and identified 104 patients with an to compare ictal phase duration and the unprovoked first seizure and a normal routine diagnostic utility of ictal duration in EEG which was followed by either an distinguishing between both seizure groups in ambulatory EEG (aEEG) or sleep derprived EEG 138 patients with either psychogenic non- (sdEEG). Patients were followed for 1 year. epileptic seizures (PNES) or epileptic seizures Sensitivities for sdEEG and aEEG were 45% (ES). Mean ictal phase duration was (specificity 91%) and 63% (specificity 95%), significantly greater in PNES (148.7s, 95% CI: respectively. Interictal epileptiform discharges 115.2-191.8) compared with ES patients on the subsequent EEG's were found an (47.7s, 95% CI: 37.6-60.6), with higher odds independent risk factor for seizure recurrence (odds ratio: 23.8, 95% CI: 7.9-71.3) of being within a year (Relative risk of 1.5). The authors PNES (24 times higher) when ictal phase was concluded that sdEEG and aEEG have similar greater than or equal to 5 minutes. Receiver diagnostic accuracies. [21] operating characteristics (ROC) curves A retrospective study assessed sleep-deprived examined the diagnostic accuracy of ictal EEG compared with standard EEG in people duration in distinguishing between seizure with suspected epilepsy (n=237). All groups producing an area under the curve of participants underwent both EEG methods, 0.80 (95% CI 0.73-0.88). A cut off value of and 69 participants had a final diagnosis of 123.5 seconds was identified as the "optimal epilepsy. Seventeen of the patients with threshold" in diagnosing PNES, with sensitivity epilepsy showed interictal epileptiform and specificity of 65% and 93% respectively. patterns in EEGs after sleep deprivation [24] (sensitivity of 25%). Sensitivity of EEG after A retrospective study with patients (n=64) with sleep deprivation was higher in patients with either localization-related epilepsy (LRE) and primary generalised epilepsies compared with genetic generalised epilepsy (GGE) and non- patients with focal epilepsies. However, EEG epileptic attacks (NEAs) were examined on the after sleep deprivation was not more sensitive occurrence of index-finger pointing (IFP) during than a subsequent repeated standard EEG, generalised convulsions by video-EEG analysis which was done in a subgroup of 55 patients. to aid in epilepsy classification. The [22] investigators found that IFP significantly occurred more in generalised tonic-clonic Use of video EEG seizures in epilepsy patients than in convulsive A study assessed standard compared with NEAs (83.6% versus 12.0%; p<0.001) and was long-term video EEG in patients older than 65 also significantly more frequent in LRE years with suspected non-convulsive seizures compared to GGE patients (96% versus 56.6%; (n=43). All participants underwent long-term p<=0.001). There were no differences between video EEG (time period not defined in the the occurrence of IFP ipsilateral, contralateral, abstract), and the first 20 minutes of each was bilateral in LRE or GGE subjects and the mean used at the standard EEG. Epileptiform angle of IFP at the metacarpophalangeal joint discharges were detected on standard EEG in was found to be 35.8 degree (SD 22.0 degree) significantly fewer people than on long-term and 3.0 degree (SD 7.2 degree) in "pointers" EEG. Non-convulsive seizures were recorded in and "nonpointers" respectively. [25] 1 person by standard EEG and in 4 people on long-term EEG. The median time to occurrence A retrospective study with patients (n=62) with of the first epileptiform activities was 46.5 either genetic generalised epilepsy (GGE), minutes. Epileptiform activity occurred during localization-related epilepsy (LRE), and non-

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 13 of 81 epileptic attacks (NEA) were assessed on the of 5 days or more was associated with an occurrence of hand postures during increased risk of the stay being inconclusive. generalised convulsions by video-EEG analysis Although the optimum cut off was 5.5 days, it to aid in differential diagnosis and epilepsy did not predict outcomes well (AUC 0.52, classification. The investigators found that LRE sensitivity 0.55, specificity 0.5). [28] patients mostly demonstrated index-finger A study assessed routine EEG (n=335) pointing (96%), GGE patients mostly compared with 3-hour outpatient video EEG demonstrated fanning at onset (91.3%) and (n=281) and 48-hour inpatient video EEG NEA patients displayed flaccid hand posture (n=247). The 3-hour outpatient video EEG (56%). Epileptic seizures significantly more recorded significantly more seizure events than often displayed fisting, fanning and index- routine EEG, and the 48-hour inpatient video finger pointing hand postures compared with EEG had a significantly higher diagnostic yield NEA (74.0% versus 32.0%, p = 0.0003; 60.3% than the 3-hour outpatient EEG. The time to versus 20.0%, p = 0.0005; 83.6% versus 12.0%, the first epileptiform discharge was shorter for p < 0.0001). NEA significantly more often patients with generalised epilepsy than for displayed flaccid posture compared with those with localised epilepsy. [29] epileptic seizures (56.0% versus 15.1%, p = 0.0001) and NEA patients were the only A retrospective analysis assessed the yield of subgroup to display claw hand posture during non-elective video EEG compared with elective NEA. [26] video EEG in adults aged 18 to 92 years (n=304). The diagnostic yield was similar – 66% A study assessed 24-hour video EEG compared for non-elective video EEG and 69% for with routine EEG for detecting epilepsy in elective video EEG; however the abstract did children (number of participants not reported not report statistical analysis of this result. in the abstract). Clinical diagnosis of epilepsy Non-elective video EEG included significantly with a minimum of 1 year of follow-up was the fewer patients with known epilepsy, the gold standard. When the video EEG did not session duration was significantly shorter, and capture any events, the sensitivity was 54% seizures and interictal epileptiform discharges and specificity was 88%. The sensitivity of were recorded less frequently compared with video EEG was higher than that of routine EEG elective video EEG. [30] (values not reported in the abstract), but specificity was similar. [27] A study assessed the utility of dynamic compared with conventional video EEG in A retrospective study assessed the use of children with epilepsy (n=200). The authors did prolonged video EEG in people with suspected not define how dynamic video EEG differed epileptic seizures (n=446) compared with from conventional video EEG. There were no suspected psychogenic non-epileptic seizures significant differences in the course of disease, (n=150). Receiver operating characteristic seizure frequency and age between the two (ROC) curves to determine optimal cut off groups. The detection rate of epileptiform points for length of stay based on futility were discharges was significantly higher with calculated. People admitted for suspected dynamic video EEG than with conventional psychogenic non-epileptic seizure were video EEG. The accuracy and specificity of significantly more likely to have an inconclusive monitoring in the V-EEG were significantly admission. There was no significant difference higher than in the routine monitoring group in the likelihood of having an inconclusive (P<0.01). Seizure frequency and number of admission if monitoring was continued for any epilepsy attacks in patients in the V-EEG group duration in patients with epileptic seizures were significantly lower than in the routine (area under curve [AUC] 0.46). For patients monitoring group (P<0.01). [31] with psychogenic non-epileptic seizures, stay

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 14 of 81 Variations of EEG A study assessed interictal EEG in children A study assessed conventional EEG diagnosed with partial epilepsies (n=35) administered by a registered technician compared with children in whom the diagnosis compared with StatNet electrode EEG of epilepsy was excluded (n=35). The authors administered by a ‘trained epilepsy fellow’ for aimed to develop a multivariable diagnostic identification of non-convulsive status prediction model based on EEG functional epilepticus (n=17). Each patient received both network findings. Periods of resting-state EEG, methods of EEG. The inter-observer free of abnormal slowing or epileptiform agreement for detection of abnormal findings activity, were selected to construct functional was 0.83 for StatNet and 0.75 for conventional networks of correlated activity. Multiple EEG. The StatNet EEG took less time to arrive network characteristics previously used in and to set up than conventional EEG. There functional network epilepsy studies were was no significant difference in the ‘percentage calculated used to build a decision tree based, of artifact duration’ between the EEG methods. prediction model. Based on epileptiform EEG [32] activity only, EEG results supported the diagnosis of epilepsy with a sensitivity of 77% A study assessed prolonged ambulatory EEG in and specificity of 91%. The prediction model people who had undergone a routine EEG had a sensitivity of 96% and specificity of 95% immediately before the ambulatory EEG in correctly differentiating patients from (n=72). The median duration of ambulatory controls (AUC 89%). [35] EEG was 22.5 hours. The sensitivity of ambulatory EEG was more than double that of A study assessed the yield of 24–72 hour routine EEG when a positive test was defined ambulatory EEG in people aged 60 years and as epileptiform discharges. However, the older (n=156). Potentially diagnostic findings sensitivity between the two types of EEG was were seen in 37% of patients, which were not significantly different if a positive test was classed as epileptiform discharges, an epileptic defined included non-epileptiform seizure, or a typical non-epileptic event. Focal abnormalities. The tests showed no significant slowing on routine EEG predicted epileptiform differences in specificity. Epileptic seizures abnormalities on ambulatory EEG. Longer were recorded in 26% of the ambulatory EEGs, duration of ambulatory EEG was associated but in 0 of the routine EEGs. [33] with increased likelihood of capturing a typical non-epileptic event. Age, the presence of a A retrospective study assessed a reduced focal lesion on MRI, and duration of electrode array (7 electrodes) compared with ambulatory EEG did not predict epileptiform the standard array of 10–20 electrodes. abnormalities on ambulatory EEG. [36] Overall, clips from 100 recordings (50 ictal and 50 non-ictal) were selected and the electrodes A UK-based safety study assessed photic were reduced digitally before blind review by 2 stimulation during EEG in adults and children specialists. For the detection of any seizure, (n=5,383). Photic stimulation elicited the reduced array EEG had a sensitivity of 70% generalised photoparoxysmal responses in 79 and specificity of 96%. Sensitivity for patients (1.5%) who had no generalised identifying encephalopathic patterns was 62% epileptiform discharges during standard EEG. and specificity was 86%. Focal seizures were Seizures were provoked in 39 patients (0.7%), more readily identified by the reduced array including 2 generalised tonic-clonic seizures. than were generalised ictal patterns. The Non-epileptic attacks were provoked in 49 authors concluded that the reduced electrode patients (0.9%). Photic stimulation yielded array was insufficiently sensitive for seizure potentially useful information in 3.1% of detection. [34] patients, and in 2.3%, photic stimulation

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 15 of 81 provided the only useful information in the the emergency department physician and EEG. [37] neurologist were asked clinical questions about the patient. The sensitivity and specificity of A study assessed the effects of ED physicians' diagnosis of the presence of hyperventilation as an EEG activation method seizure were both 88%. The inter-observer in 22 patients with focal epilepsy and 22 reliability for emergency department control participants. Standard low-resolution physicians and neurologists was reported to be electromagnetic tomography was used to moderate. Patients with abnormal EEG results analyse the neural generators of the EEG signal were prescribed new medication and changes and computed EEG lagged coherence, an index in therapy significantly more often than for of functional connectivity, between 19 regions patients with normal results. [40] of interest, and built a weighted graph for each band in every subject, and characteristic path Salzburg criteria length and clustering coefficients were calculated. Hyperventilation significantly A retrospective study assessed the diagnostic increased EEG neural generators; particularly in accuracy of the EEG criteria for diagnosis of the cingulate cortex. Functional connectivity non-convulsive status epilepticus proposed by was increased by hyperventilation in delta, a panel of experts at the fourth London- theta, alpha, and beta bands in the epileptic Innsbruck Colloquium on Status Epilepticus in group but only in theta band in the control Salzburg in 2013 (Salzburg criteria). EEG group. Intergroup analysis of mean lagged recordings from patients admitted for coherence, clustering coefficients and neurological symptoms or signs to three characteristic path length showed significant centres in Denmark and Austria. Participants differences for group, condition and band. aged 4 months and older were included from Analysis of variance for characteristic path the centres in Denmark, and were older than length also showed significant interactions for 18 years in the Austrian centre. Participants group x condition and group x band. [38] were sorted into two groups: consecutive patients under clinical suspicion of having non- EEG in the emergency department convulsive status epilepticus (the clinical validation group, n=120) or consecutive A retrospective study assessed EEG in the patients with abnormal EEG findings but no emergency department after a first seizure in clinical suspicion of non-convulsive status patients aged 17 years or older who were epilepticus (the control group, n=100). Two candidates for discharge without anti-epileptic raters blinded to all other patient data drug treatment (n=71). Patients who had retrospectively analysed the EEG recordings laboratory tests or neuroimaging that appeared and, using the Salzburg criteria, categorised to identify the cause of the seizure were patients as in non-convulsive status epilepticus excluded. A 30-minute EEG was performed in or not in non-convulsive status epilepticus. The the emergency department by an EEG reference standard was inferred from all technician. EEGs were read by a second, clinical and para-clinical data, therapeutic blinded specialist. Overall, 24% of the patients response, and the final outcome. According to had a diagnosis of epilepsy. Inter-rater the reference standard, 36% of patients in the agreement for EEG interpretation was 69%. validation group had non-convulsive status [39] epilepticus. In the validation cohort, sensitivity A prospective study assessed clinical and EEG of the Salzburg criteria was 97.7%, specificity findings in people admitted to an emergency was 89.6%, and overall accuracy was 92.5%. department with seizure or seizure-like Positive predictive value was 84.0% and symptoms (n=110). EEG was advised for all negative predictive value was 98.6%. Three patients after an initial evaluation. Before EEG, people in the control group fulfilled the

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 16 of 81 Salzburg criteria and were therefore false Neuroimaging positives (specificity 97.0%, sensitivity not An observational study assessed early calculable). Inter-rater agreement was high for comprehensive care compared with standard both the Salzburg criteria (87%) and for the care in people presenting to the emergency reference standard (95%). Therapeutic changes department with a first unprovoked seizure occurred significantly more often in the group (n=183). Standard care (n=70) involved of patients fulfilling Salzburg criteria than in consultation in the emergency department, those who did not. [41] EEG and CT. Early comprehensive care (n=113) A retrospective study assessed the Salzburg involved an epilepsy specialist consultation in Criteria for diagnosis of non-convulsive status the emergency department, routine or long- epilepticus. EEGs of 50 non-hypoxic patients term EEG, MRI, and 3 follow-up consultations. with diagnosis of non-convulsive status Long-term EEG and MRI investigations were epilepticus at discharge and 50 controls with performed significantly more often in people in abnormal EEGs. In patients without pre- the early comprehensive care group. A final existing epileptic encephalopathy, the diagnosis was obtained in significantly more following criteria were applied: more than 25 people in the early comprehensive care group epileptiform discharges per 10-second epoch than in the standard care group. Significantly (that is more than 2.5 per second) and patients more people in the early comprehensive care with epileptiform discharges less than or equal group attended a 3-month follow-up to 2.5 per second or rhythmic delta/theta appointment than in the standard care group. activity exceeding 0.5/s and at least one At 12 month follow-up, the time to first additional criteria. The additional criteria were recurrence was significantly longer in the early clinical and EEG improvements from comprehensive care group. [43] antiepileptic drugs, subtle clinical phenomena, or typical spatiotemporal evolution. In cases of MRI strength fluctuation without evolution or EEG A study assessed MRI of differing strengths improvement without clinical improvement, (1.5 tesla and 3 tesla) in patients who received possible non-convulsive status epilepticus was MRI of both strengths (n=804). Scans at 1.5 diagnosed. For identification of rhythmic tesla were conducted in 1995–2004 and scans delta/theta activity, the following criteria were at 3 teals were conducted in 2004–2011. Most compared: (test condition A) continuous delta- participants had focal epilepsy. On the second theta activity without further rules, (B) scan, 37% were normal and 20% showed American Clinical Neurophysiology Society's incidental findings. Positive findings included (ACNS) standardised critical care EEG hippocampal sclerosis (13%), malformations of terminology criterion for rhythmic delta cortical development (8%), other abnormalities activity, and (C) ACNS criteria for rhythmic (4%) and previous surgery (18%). A total of 37 delta activity and fluctuation. False positive (5%) relevant new diagnoses were made on the rates in controls dropped significantly from 3T scans not previously seen at 1.5T. [44] 28% (condition A) to 2% (condition B) and to 0% (condition C). Application of test condition PET-MRI versus PET-CT C in the group with non-convulsive status A study assessed PET-MRI compared with epilepticus gave one false negative (2%). PET-CT for identifying the focus of seizures in Possible non-convulsive status epilepticus was children aged 2–19 years with localised diagnosed based on fluctuations in 57.1% and epilepsy (n=35). All imaging was independently EEG improvement without clinical reviewed by 5 readers. The image quality did improvement in 14.2%. [42] not differ significantly between the two methods. The accuracy of PET-MRI was not

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 17 of 81 inferior to that of PET-CT for localisation of a monitoring unit, ammonia levels were checked seizure focus. [45] at baseline, within 1 hour of the event and in some patients 24 hours after the event. The EEG versus MRI investigators reported that "the change in A retrospective diagnostic study assessed EEG ammonia postictally from baseline was and MRI in children aged 1 month to 18 years significantly different among the three groups who had a first non-febrile seizure (n=248, (p=0.004). The area under the receiver 63% generalised, 36% focal). EEG results were operator characteristic (ROC) curve for pathological in 35% of participants, and MRI postictal ammonia to distinguish GCS from showed probable epileptogenic lesions in 23%. other groups was 0.88 (95% confidence In the following 48 months, almost a third of interval [CI] 0.69-0.96) suggesting ammonia be children had further seizures, which were a good test for differentiating epileptic GCS significantly related to EEG findings. EEG had from other events An ammonia level of >=80 sensitivity of 60%, specificity of 78% and mumol/L correctly classified 80% of our positive predictive value of 52%. MRI had patients (sensitivity 53.9%, specificity 100%). sensitivity of 36%, specificity of 74% and [48] positive predictive value of 34%. [46] A cross-sectional study undertook venous blood gas analysis in individuals (n=94) who Other tests visited the emergency department after seizure episodes. Venous blood gas analysis 1-hour Biomarkers post seizure found values of: pH<7.245 A retrospective case-control study assessed [sensitivity 80% (95%CI: 44-96), negative markers to predict diagnosis of occipital lobe predictive value 96.9% (95%CI: 88.3-99.4)], epilepsy (OLE). 19 patients with OLE and 57 bicarbonate<17.1 mmol/L [sensitivity 80% with temporal lobe epilepsy (TLE) were (95%CI: 44-96), negative predictive value 97% identified; three sequential case-control (95%CI: 89-99.5)], base excess<-11.1 mEq/L patients with TLE were matched with each [sensitivity 80% (95%CI: 44-96), negative patient with OLE. Visual symptoms occurred in predictive value 97% (95%CI: 89-99)], and patients with OLE (8/19) but not patients with lactate>7.65 mmol/L [sensitivity 80% (95%CI: TLE (P < 0.0001). Occipital interictal spikes (IIS) 44-96), negative predictive value 96.6% occurred in only 6 patients with OLE (P < (95%CI: 87-99)]. The investigators concluded 0.0001). IIS in the posterior temporal lobe that venous blood gas analysis for pH, base occurred in five with OLE and one with TLE (P excess, lactate and bicarbonate immediately = 0.003). IIS involved more than one lobe of one hour after the last epileptic seizure the brain occurred in 11 patients with OLE episode, could be used to predict whether the (11/19) and 9 with TLE group. (P = 0.0003) patient will have a seizure recurrence in the Multilobar resection was needed in15 patients emergency department. [49] with OLE but in only one with TLE (P < A retrospective cohort study with individuals 0.0001). [47] with generalised tonic-clonic seizures (n=195) A prospective cohort study with epilepsy and patients with other seizures (n=106) patients (n=78) examined whether postictal including syncopes (n-52) psychogenic non- transient hyperammonemia can distinguish epileptic seizures (n=17) and complex focal between different seizure types. Patients were seizures (n=37) assessed the value of serum classified according to seizure type: generalised lactate as a diagnostic biomarker of 'transient convulsive seizures (GCS), psychogenic loss of consciousness'. Blood samples were nonepileptic seizures with convulsions (PNES- taken within 2 hours of the event and serum C), or focal seizures (FS). In an epilepsy lactate levels were significantly higher in

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 18 of 81 patients with generalised tonic-clonic seizures were higher in the study group compared with p<0.001) compared to other seizure types. The control group. Ischemia-modified area under the receiver operator characteristic albumin/albumin ratios in study and control (ROC) curve was 0.94 (95% CI 0.91-0.96). For groups were 1555.3 IU/g and 462.4 IU/g (P < a cut-off concentration of 2.45mmol/l, the .001) initially and 1431.4 IU/g and 383.6 IU/g sensitivity was 0.88 and the specificity 0.87. (P < .001) at follow up, respectively. [53] [50] The use of serum metalloproteinase- 3 (MMP- A prospective cohort study with individuals 3) levels as a biomarker for diagnosis of (n=85) presenting at emergency department, epilepsy was assessed in 227 individuals with with generalised tonic-clonic seizures (GTCS) epilepsy and 97 healthy control individuals. or syncope assessed serum concentrations of Individuals with epilepsy had significantly creatine kinase (CK) and lactate as biomarkers lower serum MMP-3 levels than those in the to distinguish between different epilepsies. control group. Levels of serum MMP-3 were serum lactate levels were significantly greater significantly greater in males compared to in GTCS patients compared to syncope females, and in both epilepsy and control patients whilst there were no differences subjects, were strongly correlated with age. between patient groups with CK. Serum lactate For age 20-40 years, cut-off values for MMP-3 was highly sensitive and specific at a cut off of in 23.87ng/ml and 12.31ng/ml in males and 2.45mmol/l at 1 hour post event in females respectively, the sensitivity and distinguishing GTCS as a causal factor for specificity for patients with epilepsy versus "impairment of consciousness" from syncope controls were 72.22% and 76.67% for males, with high sensitivity (0.94) and specificity and 45% and 94.12% for females. For age (0.93), comparable results were found for >=40 years with cut-off MMP-3 second hour post seizure. [51] concentrations of 20.70ng/ml and 10.92ng/ml, sensitivity and specificity were 85.71% and Ictal-interictal continuum (IIC) continuous EEG 47.62% versus 85.62% and 100% for male and (cEEG) patterns are associated with poor female groups, respectively. The investigators outcomes in epilepsy and may represent concluded that MMP-3 as a diagnostic tool in "electroclinical" status epilepticus. In a cross epilepsy is dependent on sex of patients and is sectional study with hospitalised individuals limited to certain age brackets. [54] (n=18) FDG-PET imaging was evaluated as a complementary metabolic biomarker of status A study assessed the utility of serum epilepticus among patients with IIC cEEG metalloproteinase- 2 (MMP-2) levels as a patterns. FDG-PET hypermetabolism was biomarker for diagnosis of epilepsy in 233 common (61%) and significantly predicted individuals diagnosed with epilepsy or 97 electrographic and electroclinical status health control subjects. Individuals with epilepticus; sensitivity 79% [95% CI 53-93%], epilepsy had significantly lower serum MMP-2 and specificity 100% [95% CI 51-100 %] concentrations compared with control p=0.01. [52] subjects. In both subjects with epilepsy and control subjects, age was significantly Patients attending emergency department due correlated with serum MMP-2 levels and no to seizures (n=40) or control subjects (n=40) significant gender differences were observed. were assessed on arrival and four hours later to A concentration cut off value of 175.40ng/ml determine levels of ischemia-modified albumin produced a sensitivity of 71.13% and a (IMA) as a biomarker for differential diagnosis specificity of 62.66% in differentiating epilepsy of seizure. Mean levels of IMA were 61.5 subjects from control. [55] IU/mL and 18.5 IU/mL (P < .001) at initial assessment and 56.7 IU/mL and 15.4 IU/mL (P A retrospective study with individuals (n=270) < .001) when followed up at 4 hours, levels diagnosed with either generalised tonic-clonic

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 19 of 81 seizures (GTCS), psychogenic nonepileptic 0.705-1.095) and negative likelihood ratio of seizures (PNES) or syncope, assessed serum 1.092 (95% CI 0.941-1.268). Comparing lactate as a diagnostic biomarker for seizure groups (epileptic seizures versus NEEs; distinguishing between different epilepsy ES versus syncope; ES versus PNEEs) resulted events. Serum lactate levels were significantly in non-significant positive likelihood ratio for increased in GTCS patients compared with pooled accuracy measures for urinary PNES and syncope patients. Serum lactate incontinence. The authors concluded that levels in female GTCS patients were urinary incontinence provided no benefit as a significantly greater than in males, however the diagnostic aid in distinguishing between ES and suggested cut-off value of 2.43mmol/l whilst NEEs. [58] providing a high specificity for both gender groups produced a sensitivity of 0.85 and 0.64 Accelerometer-based diagnosis for males and females respectively, therefore A study assessed accelerometer-based time- such cut off point "might not have a frequency mapping for differentiating between discriminative effect between GTCS, PNES, epileptic seizures and psychogenic non- and syncope in female patients". [56] epileptic seizures (n=35; 56 events). Twenty- A study assessed 90 patients with epilepsy and six patients had psychogenic non-epileptic control subjects (unclear on total number of seizures, eight had epileptic seizures, and one subjects in each arm) to determine expression had both seizure types. The time-frequency of four epilepsy-associated microRNAs (miR- maps were derived from fast Fourier 106b, miR-146a, miR-194-5p and miR-301a) transformations to determine the dominant by serum analysis. Three microRNAs (miR- frequency for sequential blocks for the course 106b, miR-146a and miR-301a) were of each event. The coefficient of variation of significantly increased; miR-194-5p was limb movement frequency was less for significantly lowered in epilepsy subjects psychogenic non-epileptic seizures than for the compared with control. Both miR-106b and epileptic seizure events. A blinded review of miR-146a were correlated with NHS3 score in the time-frequency maps by an epilepsy epilepsy subjects and receiver operating specialist was accurate in differentiating characteristic (ROC) value for miR-106b was between the event types, that is, 38 of 41 non- the highest result in predicting epilepsy epileptic (92.7%) and 8 of 6 epileptic seizures (AUC=0.786) compared to serum miR-146a (75%), were diagnosed correctly, with 7 events (AUC=0.774), miR-194 (AUC=0.686) or miR- classified as non-diagnostic. Using a coefficient 310a (AUC=0.696). Combination of both miR- of variance cut-off score of 32% resulted in 106b and miR-146a improved similar classification accuracy, with 42 of 45 sensitivity/specificity for prediction of epilepsy non-epileptic (93%) and 10 of 11 epileptic (AUC=0.887). [57] seizure (91%) events correctly diagnosed. [59]

Signs and symptoms of epilepsy Electromyography A systematic review including 5 studies A study assessed electromyography using a focusing on the presence of urinary wearable device for the automatic detection of incontinence in epileptic seizures (ES) and non- generalised tonic-clonic seizures in people epileptic events (NEE) such as psychogenic referred for long-term video-EEG monitoring nonepileptic events (PNEEs) and syncope, as a on generalised tonic-clonic seizures (n=71). diagnostic aid in distinguishing between ES and Seizure detection was real-time and fully NEEs. The investigators found pooled accuracy automated. The reference standard was the measures of urinary incontinence (ES versus evaluation of video-EEG recordings by trained NEEs) yielded sensitivity 38%, specificity 57%, experts, who were blinded to data from the positive likelihood ratio of 0.879 (95% CI device. Reading the seizure logs from the

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 20 of 81 device was done blinded to all other data. The A prospective study assessed the effects of mean recording time per patient was 53 hours investigation of the cause of first seizures in and the total recording time was 3736 hours, people referred to a first seizure clinic (n=200) and device deficiency time was 193 hours compared with a historical cohort. Most (4.9% of the total time the device was turned referrals to the first seizure clinic were from on). The sensitivity of the wearable device was general practitioners and emergency 93.8% (30 of 32 generalised tonic-clonic departments. The mean waiting time for first seizures were detected). Median seizure assessment was significantly shorter for the detection latency was 9 seconds, and the false first seizure clinic cohort than for the historical alarm rate was 0.67 per day. [60] cohort. A diagnosis was established at first- contact in 80.5% of cases while 16.0% of Diagnosis after a first seizure patients needed a second visit. Eighty-two A prospective observational study assessed patients (41.0%) were diagnosed with epilepsy. diagnosis of idiopathic partial epilepsies after An abnormal EEG was found in 93.9% of the first unprovoked focal seizure in children patients diagnosed with epilepsy. Sixty-three aged 1 month to 17 years (n=107) who were patients were started on anti-epileptic drugs. In followed-up for at least 5 years. A specific 18% of cases, driving restrictions were syndrome (childhood epilepsy with centro- initiated. The most common non-seizure temporal spikes, Panayiotopoulos syndrome diagnosis was syncope (24.0%). [63] and, idiopathic childhood occipital epilepsy of Gastaut) was diagnosed in 75% of children Motion capture systems after a first seizure, and 90% of children had a A prospective study assessed a single-camera diagnosis at the end of the 5-year follow-up. In 3D video monitoring system (using the the 27 children who had no initial classification, Microsoft Kinect camera) compared with a 6- 11 patients (41%) had no change in diagnosis, camera motion-capture system, both used over and 16 patients (59%) were diagnosed with a the same bed in the epilepsy monitoring unit. specific syndrome or atypical epilepsy. [61] The correlation between the single camera system and the 6-camera system was 84.2%. A retrospective study assessed the yield of Movements of interest (n=42) arising from diagnostic investigations in people referred to temporal (n=19) and extra-temporal (n=23) an epilepsy clinic after a first seizure (n=219). brain regions were analysed. The movements No diagnosis of epilepsy was made in 38 of interest showed significant differences in people (17%), with these diagnoses mainly extent and displacement of the movements being reflex syncope and psychogenic non- between the temporal and extra-temporal epileptic seizures. Of the remaining 181 people seizure types. Movements of interest who presented with seizures, 71 had evidence originating from the temporal region were of previous seizures. Of the 110 patients with significantly shorter, and had higher jerking true first seizures, 19 were provoked, most levels, compared with extra-temporal regions. frequently by drugs or alcohol. The emergency The 3D approach was faster in extracting body department had sensitivity of 74% and motion trajectories. [64] specificity of 32% for seizure diagnosis. In the true first seizure patients, EEG demonstrated Magnetoencephalography epileptiform discharges in 22 patients (21%). CT had a 16% probability of finding a A prospective study assessed the additional potentially epileptogenic structural abnormality yield of magnetoencephalography in people and MRI had 20% probability of epileptogenic who had at least 3 normal EEGs, including findings. [62] sleep EEG (n=52). The reference standard was diagnosis based on the medical records after at least 1 year of follow-up.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 21 of 81 Magnetoencephalography and EEG were (n=20). Overall, 150 video clips from 50 performed simultaneously for 1 hour. Epilepsy patients who each had 3 seizures of the same was diagnosed in 22 patients, and type were reviewed. The videos included 37 magnetoencephalography results supported series of epileptic seizures, 8 series of this diagnosis in 9 people. The overall physiological events, and 5 series of sensitivity of magnetoencephalography was psychogenic non-epileptic seizures. The 41%, and the additional sensitivity over EEG diagnostic accuracy for epileptic seizures and alone was 18%. The specificity of non-epileptic events increased significantly magnetoencephalography was 93%, with from 61% to 66% after viewing all 3 seizures. normal magnetoencephalography in 28 of 30 Epilepsy specialists had sensitivity of 77% and patients without a diagnosis of epilepsy. [65] specificity of 81%; neurologists had sensitivity of 66% and specificity of 70%; and intern A study [Jin2017] assessed resting-state physicians had sensitivity of 59% and functional connectivity of specificity of 66%. A wide range of diagnostic magnetoencephalography for differentiating accuracy was found across the various seizures between mesial temporal lobe epilepsy (n=46) types – 90% for generalised tonic-clonic and healthy controls (n=46) and between right seizures; 80% for dialeptic seizures; 76% for and left mesial temporal lobe epilepsy (n=23 automotor seizures; 53% for myoclonic for each subtype), with use of machine learning seizures; 48% for hypermotor seizures; 44% algorithms. The optimum model identified for gelastic or dacrystic seizures; and 43% for people with mesial temporal lobe epilepsy with psychogenic non-epileptic seizures. [68] sensitivity of 96% and specificity of 94%, and differentiated left from right mesial temporal A study assessed physicians’ ability to lobe epilepsy with sensitivity of 96% and recognise paroxysmal neurological events specificity of 94%. [66] based on video observations, when blind to clinical data and final diagnosis. Overall, 12 Video observation-based diagnosis patients’ videos were used, 6 epileptic seizures, A study assessed the accuracy of neurologists’ 4 psychogenic non-epileptic seizures, and 2 diagnosis based on clinical history and video other non-epileptic events. If an epileptic recordings of paroxysmal events. Overall, 47 seizure was diagnosed, the physician was neurologists reviewed 12 clinical histories and asked to classify the type of epilepsy. videos. The median diagnostic accuracy for all Participating physicians (n=145) included paroxysmal events was significantly higher for neurologists (59%), neuropsychiatrists (26%), observation (75%) than for clinical history neurology residents (10%), and psychiatrists (67%). This was largely due the difference in (5%), most of whom (61%) were not EEG diagnostic accuracy in the subgroup of patients readers. The median proportion of correct with psychogenic non-epileptic seizures (83% diagnoses was 75%. Higher frequency of for observation and 67% for clinical history). exposure to psychogenic epileptic seizures was Higher diagnostic accuracy and increased inter- associated with better recognition of this rater agreement may be seen with higher levels disorder. Physicians were significantly more of training; however the abstract did not accurate in diagnosing epilepsy than include statistical analysis for these outcomes. psychogenic non-epileptic seizures. The Physicians with higher levels of training were median proportion of correct diagnoses for more confident with diagnosis based on epilepsy classification was 50%). [69] observation. [67] Neuropsychological assessment A study assessed the accuracy of seizure diagnosis by epilepsy specialists (n=8), No evidence was identified. neurologists (n=12) and intern physicians

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 22 of 81 2014 surveillance summary underlying cause of epilepsy and said that ‘there should be guidelines about who, when Two studies looked at how psychological and with what should be tested. There is also comorbidity and other factors relating to an increasing role for immune evaluation in epilepsy and its treatment affect health-related some circumstances – both of these have quality of life (HRQoL) in young people and implications for management, albeit in the adults. latter in tertiary care’. A study on Early-Life A systematic review sought to identify factors Epilepsies and the Emerging Role of Genetic that predicted HRQoL and resource use in Testing which concluded that ‘genetic adults with epilepsy. Age, gender and marital investigations, particularly broad sequencing status were not associated with HRQoL, methods, have high diagnostic yields in newly whereas education level and employment diagnosed early-life epilepsies regardless of status affected HRQoL evidence was key clinical features’ was identified. inconsistent. Depression, anxiety, seizure Initial intelligence gathering identified on-going frequency were predictive of poor HRQoL. research on Finding out the genetic cause of [70] Juvenile Myoclonic Epilepsy which aims to find A prospective cohort study assessed how the genetic cause for Juvenile Myoclonic epilepsy status and psychiatric and other Epilepsy by comparing the genetic code in comorbidities affect HRQoL in young people these patients with that in people who do not with epilepsy. Around one-quarter (25.6%) of have epilepsy using clues from their EEG or young people with epilepsy had any psychiatric brainwave test that is used to help diagnose disorder at follow-up, more than a third epilepsy. (39.0%) had any neurodevelopmental spectrum disorder. A psychiatric disorder was Impact statement significantly associated with worse HRQoL Overall, the new evidence supports the across the majority of the quality of life scales. existing recommendations. There is some [71] evidence that indicates a role for automated algorithms, and that machine learning may help Topic expert feedback with interpretation of EEGs in the future. [16- There was a mix of opinion from topic experts 19] There is also emerging evidence on whether the recommendation on concerning the role of genetic testing in the investigations should be updated: 2 experts diagnosis of epilepsy but the data are too thought it should be updated, while 1 though it limited at present to be considered within the did not need updating. One topic expert said guideline. there was more useful information from Functional MRI and PET scanning; and another New evidence is unlikely to change guideline topic expert highlighted that there is an recommendations. increasing role for genetics in diagnosing the

1.7 Classification

Recommendations derived from these review questions 1.7.1 Epileptic seizures and epilepsy syndromes in children, young people and adults should be classified using a multi-axial diagnostic scheme. The axes that should be considered are:

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 23 of 81 description of seizure (ictal phenomenology); seizure type; syndrome and aetiology. [2004] 1.7.2 The seizure type(s) and epilepsy syndrome, aetiology, and co-morbidity should be determined, because failure to classify the epilepsy syndrome correctly can lead to inappropriate treatment and persistence of seizures. [2004] 1.7.3 Children, young people and adults with epilepsy should be given information about their seizure type(s) and epilepsy syndrome, and the likely prognosis. [2004]

the new ILAE classification, changes in Classification terminology and clarification of definitions, which would have wider implications 2018 surveillance summary throughout the guideline. References were A retrospective study assessed the provided to several publications: International League Against Epilepsy (ILAE) Operational classification of seizure types by definition of epilepsy (one unprovoked seizure the International League Against Epilepsy: with a probability of further seizures). Patients Position Paper of the ILAE Commission for (n=1,006) were categorised depending on Classification and Terminology whether they had one unprovoked seizure (new definition of epilepsy; n=152) or 2 or Instruction manual for the ILAE 2017 more unprovoked seizures (traditional operational classification of seizure types definition of epilepsy; n=854) at the time of Classification of the epilepsies: New concepts epilepsy diagnosis. Patients diagnosed for discussion and debate—Special report of according to the new definition showed a the ILAE Classification Task Force of the higher proportion of subjects with an abnormal Commission for Classification and Terminology neurological examination and with focal seizures. The two samples differed in the ILAE classification of the epilepsies: Position presence of at least one of the factors paper of the ILAE Commission for predicting seizure recurrence (focal seizures or Classification and Terminology abnormal findings in at least one among the A definition and classification of status following: neurologic examination, EEG, and epilepticus--Report of the ILAE Task Force on neuroimaging. Long-term recurrence in Classification of Status Epilepticus patients diagnosed with the new definition was 83.6% at 10 years and 89.1% at 15 years. The Impact statement probability of early remission did not differ The new ILAE classification, changes in between the two groups. [72] terminology and definitions should be considered and reflected within this 2014 surveillance summary recommendation and throughout the guideline. No relevant evidence was identified.

New evidence identified that may change Topic expert feedback current recommendations. Two topic experts said that the recommendation should be updated to reflect

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 24 of 81 1.8 Management

Recommendations in this section of the guideline 1.8.1 Children, young people and adults with epilepsy should have an accessible point of contact with specialist services. [2004] 1.8.2 All children, young people and adults with epilepsy should have a comprehensive care plan that is agreed between the person, family and/or carers where appropriate, and primary care and secondary care providers. This should include lifestyle issues as well as medical issues. [2004] 1.8.3 Epilepsy specialist nurses (ESNs) should be an integral part of the network of care of children, young people and adults with epilepsy. The key roles of the ESNs are to support both epilepsy specialists and generalists, to ensure access to community and multi-agency services and to provide information, training and support to the child, young person or adult, families, carers and, in the case of children, others involved in the child's education, welfare and well-being. [2004] 1.8.4 Healthcare professionals have a responsibility to educate others about epilepsy so as to reduce the stigma associated with it. They should provide information about epilepsy to all people who come into contact with children, young people and adults with epilepsy, including school staff, social care professionals and others. [2004]

2018 surveillance summary 2014 surveillance summary A Cochrane review of 18 studies (RCTs, No relevant evidence was identified. controlled or matched trials, cohort studies or other prospective studies with a control group, Topic expert feedback and time series studies; n=NR) evaluating the No topic expert feedback was relevant to this effects of any specialised or dedicated evidence. intervention beyond usual care in adults with epilepsy, identified 16 separate interventions, Impact statement which were classified into 7 groups. It was The evidence supports the current noted that the specialist epilepsy nurse and recommendation concerning the use of self-management education had some epilepsy nurses and self-management evidence of benefit, but that most of the education in the management of epilepsy. studies had methodological weaknesses and no single model of service provision could be recommended. [73] New evidence is unlikely to change guideline recommendations.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 25 of 81 1.9 Pharmacological treatment

Recommendations in this section of the guideline Note: see appendix E for further details of pharmacological treatment. The GDG is aware of the contraindications to prescribing carbamazepine to some people of Han Chinese or Thai origin. Recommendations in this section offer alternatives, and so no specific recommendations are made for these groups. The GDG is also aware of specific issues with prescribing sodium valproate to girls and women of childbearing age. Recommendations in this section offer alternative prescribing options for this group. Recommendations 1.9.1.10, 1.9.17.3, 1.9.17.6, 1.9.17.9 and 1.15.1.4 also provide additional specific information of relevance when considering prescribing AEDs to women of childbearing age. NICE has also issued guidance on the use of retigabine as an option for the adjunctive treatment of partial (the term focal has been used in this guideline) onset seizures with or without secondary generalisation in adults aged 18 years and older with epilepsy in Retigabine for the adjunctive treatment of partial onset seizures in epilepsy (NICE technology appraisal guidance 232). The GDG is also aware of specific issues with prescribing sodium valproate to girls and women of childbearing age. In January 2015, the MHRA issued a strengthened warning stating that valproate should not be prescribed to female children, female adolescents, women of childbearing potential or pregnant women unless other treatments are ineffective or not tolerated. Recommendations in this section offer alternative prescribing options for this group. Recommendations 1.9.1.10, 1.9.17.3, 1.9.17.6, 1.9.17.9 and 1.15.1.4 also provide additional specific information of relevance when considering prescribing AEDs to women of childbearing age. 1.9.1 General information about pharmacological treatment 1.9.1.1 Information that is provided about anti-epileptic drugs (AEDs) needs to be in the context of that provided by the manufacturer, for example, indications, side effects and licence status. [2004] 1.9.1.2 The AED treatment strategy should be individualised according to the seizure type, epilepsy syndrome, co-medication and co-morbidity, the child, young person or adult's lifestyle, and the preferences of the person and their family and/or carers as appropriate (see appendix E). [2004] 1.9.1.3 The diagnosis of epilepsy needs to be critically evaluated if events continue despite an optimal dose of a first-line AED. [2004] 1.9.1.4 Consistent supply to the child, young person or adult with epilepsy of a particular manufacturer's AED preparation is recommended, unless the prescriber, in consultation with the child, young person, adult and their family and/or carers as appropriate, considers that this is not a concern. Different preparations of some AEDs may vary in bioavailability or pharmacokinetic profiles and care needs to be taken to avoid reduced effect or excessive side effects. Consult the summary of product characteristics (SPC) and British national formulary (BNF) on the bioavailability and pharmacokinetic profiles of individual AEDs, but note that these do not give information on comparing bioavailability of different generic preparations*,**. [New 2012] 1.9.1.5 It is recommended that children, young people and adults should be treated with a single AED (monotherapy) wherever possible. If the initial treatment is unsuccessful, then monotherapy using another drug can be tried. Caution is needed during the changeover period. [2004] 1.9.1.6 If an AED has failed because of adverse effects or continued seizures, a second drug should be started (which may be an alternative first-line or second-line drug) and built up

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 26 of 81 to an adequate or maximum tolerated dose and then the first drug should be tapered off slowly. [2004] 1.9.1.7 If the second drug is unhelpful, either the first or second drug may be tapered, depending on relative efficacy, side effects and how well the drugs are tolerated before starting another drug. [2004] 1.9.1.8 It is recommended that combination therapy (adjunctive or 'add-on' therapy) should only be considered when attempts at monotherapy with AEDs have not resulted in seizure freedom. If trials of combination therapy do not bring about worthwhile benefits, treatment should revert to the regimen (monotherapy or combination therapy) that has proved most acceptable to the child, young person or adult, in terms of providing the best balance between effectiveness in reducing seizure frequency and tolerability of side effects. [2004] 1.9.1.9 If using carbamazepine, offer controlled-release carbamazepine preparations. [new 2012] 1.9.1.10 When prescribing sodium valproate to women and girls of present and future childbearing potential, discuss the possible risk of malformation and neurodevelopmental impairments in an unborn child, particularly with high doses of this AED or when using as part of polytherapy. [new 2012]

* In November 2013, the MHRA issued new advice about oral anti-epileptic drugs (AEDs) and switching between different manufacturers' products of a particular drug. Following a review of the available evidence, the Commission on Human Medicines (CHM) has classified AEDs into 3 categories depending on the level of potential concerns related to switching between different manufacturers' products. Consult the MHRA advice for more information. ** Recommendations 1.1.1, 1.2.2, 1.3.3, 1.3.5 and 1.3.10 describe the principles of decision making and best practice in relation to effective and appropriate consultation between healthcare professionals and children, young people and adults with epilepsy.

mortality after a first seizure. Treatment of the General information about first seizure was associated with a significantly pharmacological treatment higher risk of adverse events compared with deferred treatment (moderate quality 2018 surveillance summary evidence). [74]

A Cochrane review of 6 RCTs/quasi-RCTs A Cochrane review of 10 RCTs (296 (2,439 participants) assessed the probability of participants) evaluating the efficacy of seizure recurrence, seizure remission, immediate-release carbamazepine (IR CBZ) mortality, and adverse effects of AEDs given versus controlled-release carbamazepine (CR immediately after first seizure unprovoked CBZ) in patients diagnosed with epilepsy epileptic compared with controls, in children concluded that ‘at present, data from trials do and adults. An individual participant meta- not confirm or refute an advantage for CR CBZ analysis was undertaken on data from 2 over IR CBZ for seizure frequency or adverse studies. Results indicate that immediate events in patients with newly diagnosed treatment is associated with a lower probability epilepsy’. The trials were of small size and poor of relapse at one year and at five years and a methodological quality. [75] higher probability of an immediate five-year remission (high quality evidence); but no A Cochrane review of 6 RCTs (125 difference between immediate treatment and participants) evaluating the efficacy and control in terms of five year remission at any tolerability of melatonin as add-on treatment time. Antiepileptic drugs did not affect overall for epilepsy found that no conclusion about

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 27 of 81 the role of melatonin in reducing seizure recommending new measures to avoid frequency or improving quality of life in people exposure of babies to valproate medicines in with epilepsy could be made as the included the womb, including: studies were of poor methodological quality 1) Valproate should be contraindicated in and did not systematically evaluate seizure pregnancy and women of childbearing frequency and adverse events. [76] potential not using effective contraception. A Cochrane review of 3 RCTs (246 2) This should be supported by a ‘Pregnancy participants) evaluating the efficacy and side Prevention Programme’. effect profile of sulthiame as monotherapy in people with epilepsy reported that it was not 3) A signed ‘acknowledgement’ or ‘consent’ possible make any conclusions on the efficacy form should be routinely used when women and safety of sulthiame as monotherapy due to are reviewed on an annual basis by a specialist. small sample size, poor methodological quality The MHRA and manufacturers have developed and lack of data on important outcome a Toolkit on the risks of valproate medicines in measures. [77] female patients which was published in February 2016 and supports informed choices 2014 surveillance summary about valproate use in girls and women. This No relevant evidence was identified. will be updated to match the proposed contraindications. Topic expert feedback and additional The footnote concerning the MHRA division of information antiepileptic drugs into 3 categories in 2013 No topic expert feedback was relevant to this was updated in November 2017. The MHRA evidence. have added the advice that ‘when evaluating Work was conducted by the NICE surveillance whether continuity of supply should be team in December 2017 to consider new maintained for category 2 or 3 drugs, consider: advice from the Medicines & Healthcare  perception by patients of differences in products Regulatory Agency (MHRA) on supply, for example differences in product sodium valproate use during pregnancy: presentations Sodium valproate is a well-established  co-morbid autism, mental health issues, or treatment for epilepsy. However, treatment learning disability.’ during pregnancy is associated with a risk of abnormal pregnancy outcomes and the MHRA Initial intelligence gathering identified that the website states that sodium valproate is guidance Retigabine for the adjunctive associated with a 10% increased risk of treatment of partial onset seizures in epilepsy congenital malformations, delays in early (July 2011) TA232 has been withdrawn development of children and approximately 3- because GlaxoSmithKline has discontinued fold increased risk of autistic spectrum disorder retigabine (Trobalt) due to limited usage. and approximately 5-fold increased risk of childhood autism compared with the general Impact statement study population. The association has been The new evidence supports existing recognised for many years; it appears to be recommendations concerning an AED dose-dependent and with increased risks when treatment strategy; and the evidence at valproate is taken with other medicines. present does not indicate a difference in The European Medicines Agency’s experts in outcomes between immediate-release versus medicines safety, the Pharmacovigilance Risk controlled-release carbamazepine in people Assessment Committee (PRAC) and MHRA are with epilepsy. Neither melatonin nor sulthiame are recommended for use in people with

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 28 of 81 epilepsy and the current evidence does not have clear warnings to alert prescribers, it is indicate that this would change. necessary to incorporate the new information on contraindications into the guideline An editorial change is required to remove recommendations. reference to retigabine; and the footnote relating to the MHRA categorisation of antiepileptic drugs needs to be updated with New evidence identified that may change the latest advice. current recommendations.

While recommendations for the use of sodium valproate in NICE guideline CG137 already

Recommendations in this section of the guideline 1.9.2 Initiation of pharmacological treatment 1.9.2.1 AED therapy should only be started once the diagnosis of epilepsy is confirmed, except in exceptional circumstances that require discussion and agreement between the prescriber, the specialist and the child, young person or adult and their family and/or carers as appropriate. [2004] 1.9.2.2 AED therapy should be initiated in adults on the recommendation of a specialist. [2004] 1.9.2.3 AED therapy in children and young people should be initiated by a specialist. [2004] 1.9.2.4 The decision to initiate AED therapy should be taken between the child, young person or adult, their family and/or carers (as appropriate) and the specialist after a full discussion of the risks and benefits of treatment. This discussion should take into account details of the person's epilepsy syndrome, prognosis and lifestyle. [2004] 1.9.2.5 Treatment with AED therapy is generally recommended after a second epileptic seizure. [2004] 1.9.2.6 When possible, choose which AED to offer on the basis of the presenting epilepsy syndrome. If the epilepsy syndrome is not clear at presentation, base the decision on the presenting seizure type(s). [new 2012] 1.9.2.7 AED therapy should be considered and discussed with children, young people and adults and their family and/or carers as appropriate after a first unprovoked seizure if:  the child, young person or adult has a neurological deficit  the EEG shows unequivocal epileptic activity  the child, young person or adult and/or their family and/or carers consider the risk of having a further seizure unacceptable  brain imaging shows a structural abnormality. [2004] 1.9.2.8 It should be recognised that some children, young people and adults (through their families and/or carers, in some instances) may choose not to take AED therapy following a full discussion of the risks and benefits. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 29 of 81 Topic expert feedback Initiation of pharmacological treatment No topic expert feedback was relevant to this 2018 surveillance summary evidence.

No relevant evidence was identified. Impact statement

2014 surveillance summary The absence of new evidence indicates no need to update this section of the guideline. No relevant evidence was identified.

No new evidence. Recommendations are unlikely to change.

Recommendations in this section of the guideline 1.9.3 Pharmacological treatment of focal seizures First-line treatment in children, young people and adults with newly diagnosed focal seizures 1.9.3.1 Offer carbamazepine or lamotrigine as first-line treatment to children, young people and adults with newly diagnosed focal seizures. [new 2012] 1.9.3.2 Levetiracetam is not cost effective at June 2011 unit costs*. Offer levetiracetam, oxcarbazepine or sodium valproate (provided the acquisition cost of levetiracetam falls to at least 50% of June 2011 value documented in the National Health Service Drug Tariff for England and Wales) if carbamazepine and lamotrigine are unsuitable or not tolerated. If the first AED tried is ineffective, offer an alternative from these five AEDs. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)**. [new 2012] 1.9.3.3 Consider adjunctive treatment if a second well-tolerated AED is ineffective (see recommendations 1.9.3.1 and 1.9.3.2). [new 2012] Adjunctive treatment in children, young people and adults with refractory focal seizures 1.9.3.4 Offer carbamazepine, clobazam†, gabapentin†, lamotrigine, levetiracetam, oxcarbazepine, sodium valproate or topiramate as adjunctive treatment to children, young people and adults with focal seizures if first-line treatments (see recommendations 1.9.3.1 and 1.9.3.2) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)**. [new 2012] 1.9.3.5 If adjunctive treatment (see recommendation 1.9.3.4) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist. Other AEDs that may be considered by the tertiary epilepsy specialist are eslicarbazepine acetate†, lacosamide, phenobarbital, phenytoin, pregabalin†, tiagabine, vigabatrin and zonisamide†. Carefully consider the risk– benefit ratio when using vigabatrin because of the risk of an irreversible effect on visual fields. [new 2012]

* Estimated cost of a 1500 mg daily dose was £2.74 at June 2011. Cost taken from the National Health Service Drug Tariff for England and Wales. ** February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. †At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 30 of 81

risk-benefit balance of vigabatrin versus Pharmacological treatment of focal carbamazepine monotherapy for epilepsy. seizures While no differences in visual field defects and visual disturbances were noted, the authors 2018 surveillance summary referenced an existing systematic review of observational studies that showed a high Focal (aka partial) seizures, newly prevalence of visual field defects with diagnosed/monotherapy vigabatrin monotherapy. [79] A Cochrane review undertook an individual A Cochrane review undertook an IPD analysis participant data (IPD) analysis. Data were incorporating a network meta-analysis (NMA). available for 595 of 1,192 eligible participants Data was available from 12,391 of 17,961 from 4 of 12 RCTs. The analysis assessed the eligible participants from 36 of 77 RCTs. The time to withdrawal, 6 and 12 month remission, analysis compared the time to withdrawal of and first seizure with carbamazepine compared allocated treatment, remission and first seizure to phenytoin, used as monotherapy in people of 10 AEDs that are currently used as with partial onset seizures (simple partial, monotherapy in children and adults with partial complex partial, or secondarily generalised onset seizures (simple partial, complex partial tonic-clonic seizures), or generalised tonic- or secondary generalised) or generalised tonic- clonic seizures, with or without other clonic seizures with or without other generalised seizure types. There was no generalised seizure types (absence, statistically significant difference between myoclonus). For individuals with partial onset carbamazepine and phenytoin for these seizures, carbamazepine and lamotrigine are outcomes, but the authors noted that CIs were suitable first-line treatments and that wide and they could not exclude the possibility levetiracetam may be a suitable alternative. For of important differences. There were more individuals with generalised tonic-clonic serious adverse events with phenytoin than seizures (with or without other generalised carbamazepine. While there was some seizure types), there was high-quality evidence evidence that people with generalised seizures supporting the use of sodium valproate as the may be less likely to withdraw early from first-line treatment and that lamotrigine and phenytoin than from carbamazepine, levetiracetam would be suitable alternative misclassification of seizure type may have first-line treatments. [80] impacted upon the results (the authors identified up to 48 individuals with A Cochrane review of IPD (data available for misclassification of seizure type, 32% of whom 1,151 of 1,239 eligible participants from 2 of 3 had generalised epilepsy). The authors RCTs) evaluated the effects of topiramate ‘recommend caution when interpreting the monotherapy versus carbamazepine results of this review, and do not recommend monotherapy for epilepsy in people with that our results alone should be used in partial-onset seizures or generalised onset choosing between carbamazepine and tonic-clonic seizures. In individuals with partial- phenytoin’. [78] onset seizures treated with carbamazepine, 12- month remission is achieved earlier and it is A Cochrane review of 5 RCTs (734 less likely to be withdrawn than in patients participants) compared the efficacy and safety treated with topiramate. While evidence of vigabatrin versus carbamazepine indicated no differences between the drugs monotherapy for epilepsy in children and and outcomes in people with generalised-onset adults. Data were insufficient to address the

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 31 of 81 tonic-clonic seizures, there were only a limited tolerability of pregabalin monotherapy in number of individuals identified with these people with epilepsy, reported on the results of seizures (15% of the sample). [81] 1 study that found that pregabalin had similar tolerability, but inferior efficacy to lamotrigine A Cochrane review of IPD available for 2,572 for newly diagnosed partial seizures. However, of 3,394 eligible participants from 9 of 13 trials this study was limited in design; and the review evaluated the effectiveness of lamotrigine authors stated that ‘data were too limited to monotherapy compared to carbamazepine draw any conclusions’. [85] monotherapy in people with partial onset seizures or generalised onset tonic-clonic A Cochrane review of IPD from 480 of 517 seizures. In individuals with partial onset eligible participants in 3 RCTs compared seizures Lamotrigine was significantly less oxcarbazepine and phenytoin monotherapy in likely to be withdrawn than carbamazepine, but participants with partial onset seizures or carbamazepine may be superior in terms of generalised tonic-clonic seizures with or seizure control. The results were less clear for without other generalised seizure types. There individuals with generalised onset seizures, was a statistically significant advantage for who made up only 12% of cases, as the study oxcarbazepine over phenytoin for time to authors were concerned that seizure type may treatment withdrawal, but insufficient have been misclassified in up to 50% of these evidence to suggest a difference between the participants. Methodological quality of the drugs for other outcomes. When considered by studies was rated as generally good. [82] seizure type, there was no significant advantage for either drug for generalised A Cochrane review of IPD from 836 of 1,455 epilepsy, however oxcarbazepine is eligible participants from 6 of 13 trials significantly less likely to be withdrawn than evaluated pair-wise monotherapy comparisons phenytoin in participants with partial onset in people with partial onset seizures or seizures, but it was not possible to determine generalised onset tonic-clonic seizures. Results whether oxcarbazepine is equivalent, superior indicated that that carbamazepine may be or inferior to phenytoin in terms of seizure more effective than phenobarbitone in terms control. [86] of seizure control and adverse events; and that for time to first seizure recurrence A Cochrane review of IPD available for 669 of phenobarbitone was better for partial seizures 1,119 eligible participants from 5 of 11 trials and carbamazepine for generalised seizures. investigated the time to withdrawal, remission The evidence was rated as low quality. [83] and first seizure of phenytoin compared to valproate when used as monotherapy in people A Cochrane review of 2 RCTs/quasi RCTs (163 with partial onset seizures or generalised tonic- participants) evaluated the efficacy, clonic seizures (with or without other effectiveness, tolerability and safety of generalised seizure types). There was no clobazam as monotherapy in people with new- overall difference between the drugs for these onset partial or generalised seizures. In outcomes, nor any statistical interaction previously-untreated children, there was no between treatment and seizure type. The benefit for clobazam compared with authors noted that ‘misclassification of seizure carbamazepine for still taking treatment type may have confounded the results of this (retention) at 12 months. In adolescents and review’’. [87] adults with neurocysticercosis, there was a small benefit of clobazam compared with A Cochrane review of IPD available for 599 phenytoin for retention at six months. [84] eligible participants from 4 of 8 trials evaluated the effectiveness of phenobarbitone A Cochrane review of 2 RCTs (753 [phenobarbital] monotherapy versus phenytoin participants) evaluating the efficacy and monotherapy in children or adults with partial

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 32 of 81 onset seizures or generalised onset tonic-clonic for adults with drug-resistant partial epilepsy in seizures. There was no significant differences the short-term, but is associated with adverse for seizure outcomes between the 2 AEDs, but effects. [92] a statistically significant clinical advantage for A Cochrane review of 4 RCTs (236 phenytoin in terms of treatment withdrawal participants) evaluating the efficacy and (the authors hypothesised that phenobarbitone tolerability of felbamate versus placebo when may lead to more adverse effects). Neither used as an add-on treatment for people with study was blinded and there was high refractory partial-onset epilepsy found that statistical heterogeneity between the studies. due to the ‘methodological deficiencies’ in the [88] studies there is ‘no reliable evidence to support the use of felbamate as an add-on therapy in Refractory Focal seizures people with refractory partial-onset epilepsy’. A Cochrane review of 2 RCTs (467 [93] participants) investigating the efficacy and safety of losigamone when used as an add-on A Cochrane review of 11 RCTs (2,125 therapy for partial epilepsy found that participants) evaluating the efficacy and losigamone reduced seizure frequency, but was tolerability of gabapentin as an add-on associated with more treatment withdrawals treatment for people with drug-resistant partial when used as an add-on therapy. The authors epilepsy undertook a meta-analysis on data noted that the trials were of short-term from 6 RCTs (1,206 participants). This analysis duration and uncertain quality. [89] found that gabapentin significantly reduced seizure frequency and showed increasing A Cochrane review of 11 RCTs (747 efficacy with increasing dose, but is associated participants) evaluating the short-term with significantly more adverse effects effectiveness of vigabatrin versus placebo in compared with placebo. The studies were people with drug-resistant partial epilepsy rated as low/unclear risk of bias and were of found that vigabatrin educes seizure short duration (there was no evidence for frequency, but is associated with side-effects, efficacy beyond three-months). [94] in particular fatigue or drowsiness. [90] A Cochrane review of 14 RCTs (1,958 A Cochrane review of 3 RCTs (1,311 participants) evaluating the effectiveness of participants) evaluated the efficacy and lamotrigine compared to placebo as an add-on tolerability of lacosamide as an add-on treatment for people with refractory partial treatment for people with drug-resistant partial epilepsy reported that lamotrigine was epilepsy. Compared with placebo, lacosamide effective in reducing seizure frequency, and was effective and well tolerated in the short was fairly well tolerated, but the trials were of term in adults. Higher doses of lacosamide relatively short duration. Note, this was an were associated with more adverse effects and update of 2010 Cochrane review, and no new withdrawal of the drug compared with lower studies have been included. [95] doses. Quality of the evidence was rated as moderate to high. [91] A Cochrane review of 11 RCTs (1,861 participants) evaluating the effectiveness of A Cochrane review of 5 RCTs (1,799 levetiracetam as an add-on treatment for participants) evaluating the efficacy and people with drug-resistant focal epilepsy found tolerability of eslicarbazepine acetate when that levetiracetam can significantly reduce used as an add-on treatment for people with focal seizure frequency in adults and children, drug-resistant partial epilepsy found moderate and is well tolerated, although non-specific to high evidence that, compared with placebo, changes in behaviour may be experienced in up eslicarbazepine acetate reduces seizure frequency when used as an add-on treatment

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 33 of 81 to 20% of children. The studies were rated as A Cochrane review of 5 RCTs (949 having low risks of bias. [96] participants) evaluating the efficacy and tolerability of zonisamide add-on treatment for A Cochrane review of 6 RCTs (2,009 people with drug-resistant partial epilepsy participants) evaluating the efficacy and found that zonisamide reduced seizure tolerability of pregabalin add-on treatment in frequency compared with placebo (evidence drug-resistant partial epilepsy reported that it had a low or unclear risk of bias). [101] is significantly more effective than placebo at achieving a 50% or greater seizure reduction 2014 surveillance summary and increasing seizure freedom; doses from 150 mg/day to 600 mg/day were efficacious, Three studies on the use of perampanel as an with increasing effectiveness at 600 mg doses adjunctive treatment for the treatment of focal (evidence rated as low/unclear risk of bias). seizures in epilepsy were identified. The [97] evidence suggested that compared with placebo, adjunctive treatment with perampanel A Cochrane review evaluating the efficacy and 4–12 mg once daily reduces seizure frequency tolerability of stiripentol as add-on treatment in people aged 12 years and older with for patients with focal refractory epilepsy who uncontrolled focal seizures, but adverse events are taking AEDs only identified 1 RCT (32 were high. [102-104] See Partial-onset children). There were non-significant seizures in epilepsy: perampanel as adjunctive reductions in seizure frequency and treatment Evidence summary: new medicine improvement in seizure freedom when add-on (ESNM) 7. stiripentol was compared with placebo. Due to small sample size, relative risk ratios with wide Topic expert feedback and additional confidence intervals and concerns about information external validity of the study the authors reported that conclusions could not be made As explained in the previous section to support the use of stiripentol as add-on (recommendation 1.9.1) concerns about the treatment for focal refractory epilepsy. [98] safety of sodium valproate treatment in women have triggered this exceptional review A Cochrane review of 6 RCTs (n=800) and potential update of the guideline. In evaluating the effects of tiagabine add-on addition a topic expert said that treatment in people with drug-resistant recommendations concerning the use of localisation-related seizures found that, sodium valproate (in relation to consent to compared with placebo, tiagabine reduces treatment) should be considered for an update. seizure frequency, but is associated with some adverse effects (overall study quality rated as It was also noted that there are several new high). [99] antiepileptic drugs since the revision in 2012 that should be considered for an update. A Cochrane review of 11 RCTs (1,401 participants) evaluating the effects of Topic experts highlighted the following studies: topiramate add-on treatment in people with Safety and tolerability of lacosamide as drug-resistant partial epilepsy found that in the adjunctive therapy for adults with partial-onset short-term, topiramate is three times more seizures: Analysis of data pooled from three effective compared with placebo in reducing randomized, double-blind, placebo-controlled seizures; there was no evidence of long-term clinical trials and Brivaracetam as adjunctive effects. It was noted that topiramate was treatment for uncontrolled partial epilepsy in associated with significantly more side-effects adults: a phase III randomized, double-blind, (moderate quality evidence). [100] placebo-controlled trial.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 34 of 81 Initial intelligence gathering identified that The publication from NIHR (National Institute there are two Cochrane review protocols for Health Research) A range of anti-epilepsy published investigating Perampanel add-on for drugs are effective as first-line treatment was drug-resistant partial epilepsy and Perampanel identified that concluded that ‘the findings monotherapy for epilepsy. There is also a support NICE recommendations to use technology appraisal scheduled on Perampanel carbamazepine or lamotrigine as first-line (Fycompa; E-2007; ER-155055-90) for Partial therapies for epilepsy with partial seizures, onset seizures with or without secondarily with levetiracetam as an alternative. Sodium generalised seizures or primary generalised valproate or lamotrigine are recommended for tonic clonic seizures in paediatric patients aged people with generalised tonic-clonic seizures, 4-12 with epilepsy – adjuvant treatment (TSID and levetiracetam is an alternative option. 9149). Impact statement Initial intelligence gathering identified that within Partial seizures in children and young Evidence identified in this surveillance review people with epilepsy: zonisamide as adjunctive support the effectiveness of sodium valproate therapy (March 2014) ESNM37 it states that in However, safety concerns about sodium October 2013, the approved licence for valproate in girls and women have suggested zonisamide (Zonegran, Eisai Limited; adjunctive the need to review the recommendations for treatment of partial seizures, with or without sodium valproate and its place in the secondary generalisation, in adults) was management of the epilepsies (see section on extended to include 'adolescents' and children recommendation 1.9.1). The ongoing SANADII aged 6 years and over (see the European public study will supply additional information and it assessment report (EPAR) for Zonegran). is expected that data will be available during Partial-onset seizures in epilepsy: zonisamide the update of the guideline. as monotherapy (April 2013) ESNM17 states The new evidence supports the current that in June 2012, the approved licence for recommendations concerning the effectiveness zonisamide (Zonegran; adjunctive treatment of of carbamazepine, lamotrigine, levetiracetam, partial seizures, with or without secondary oxcarbazepine as first-line treatments in generalisation, in adults) was extended to children, young people and adults with newly include monotherapy for treating partial-onset diagnosed focal seizures. There was also seizures, with or without secondary evidence indicating that phenytoin may be an generalisation, in adults with newly diagnosed effective first-line treatments in people with epilepsy. newly diagnosed focal seizures, as this is not On-going research, Trial comparing the currently recommended, the evidence base for effectiveness and cost effectiveness of this AED should be reviewed as part of the levetiracetam and zonisamide versus standard guideline update. Other AEDs that have been treatments for epilepsy: a comparison of assessed for effectiveness as first-line Standard and New Antiepileptic Drugs treatments in people with newly diagnosed (SANAD-II trial) was also identified. This study focal seizures, that are also not currently will provide updated effectiveness information recommended, are pregabalin, vigabatrin or for a range of AEDs. A topic expert also stated clobazam, however there was only limited that the recommendations for levetiracetam evidence of effectiveness of these AEDs and as needed review in terms of cost-effectiveness. such it is unlikely that this evidence would It has been noted that the current price of have an impact on the current levetiracetam is approximately half of the price recommendation. in 2011, but as it is a category M in the Drug There are several new AEDs (zonisamide, Tariff, prices can vary. losigamone, brivaracetam and perampanel)

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 35 of 81 with evidence of effectiveness in the treatment adjunctive treatment for focal seizures could of focal seizures; and indications that there are be considered in an update, however additional AEDs that might be considered as losigamone is not currently available in the UK. first-line treatment (e.g. phenytoin). Evidence concerning the use of perampanel as ESNM17 states that zonisamide could offer an a first-line and adjunctive treatment should be alternative to other anti-epileptic drugs that considered during the update of this section. are currently recommended as first-line Currently within NICE guideline CG137 when treatment in people with newly diagnosed zonisamide is recommended, the footnote focal seizures ‘because of its different states that ‘at the time of publication (January mechanism of action, once-daily dosing, and 2012), this drug did not have UK marketing adverse event and interaction profiles (unlike authorisation for this indication and/or some other anti-epileptic drugs, zonisamide is population (see appendix E for details). not thought to affect the pharmacokinetics of Informed consent should be obtained and other medicines, such as oral contraceptives, documented.’ Appendix E provides information through cytochrome P450-mediated on each drug; for zonisamide: ‘At the time of mechanisms)’. Evidence concerning the use of publication, zonisamide did not have UK zonisamide monotherapy for focal seizures marketing authorisation for use in children should be considered in the update. younger than 18 years owing to insufficient The new evidence supports the current data on safety and efficacy (SPC).’ This is now recommendations concerning the use of not the case, at least for adjunctive treatment vigabatrin, lacosamide, eslicarbazepine acetate, of partial seizures, with or without secondary gabapentin, lamotrigine, levetiracetam, generalisation; hence it is recommended that pregabalin, tiagabine, topiramate and the footnote for recommendation 1.9.3.5 zonisamide as adjunctive treatments for should be removed and reference to ESNM37 refractory focal seizures. The guideline does added (editorial correction). The change in not recommend felbamate or stiripentol as licencing does not change the recommendation adjunctive treatments and current evidence content. does not indicate that this would change. The evidence concerning the effectiveness and New evidence identified that may change safety profile of losigamone and brivaracetam current recommendations. (which are not currently recommended) as an

Recommendations in this section of the guideline 1.9.4 Pharmacological treatment of newly diagnosed generalised tonic–clonic (GTC) seizures First-line treatment in children, young people and adults with newly diagnosed GTC seizures 1.9.4.1 Offer sodium valproate as first-line treatment to children, young people and adults with newly diagnosed GTC seizures. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.4.2 Offer lamotrigine if sodium valproate is unsuitable. If the person has myoclonic seizures or is suspected of having juvenile myoclonic epilepsy (JME), be aware that lamotrigine may exacerbate myoclonic seizures. [new 2012] 1.9.4.3 Consider carbamazepine and oxcarbazepine** but be aware of the risk of exacerbating myoclonic or absence seizures. [new 2012] Adjunctive treatment in children, young people and adults with GTC seizures

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 36 of 81 1.9.4.4 Offer clobazam**, lamotrigine, levetiracetam, sodium valproate or topiramate as adjunctive treatment to children, young people and adults with GTC seizures if first-line treatments (see recommendations 1.9.4.1, 1.9.4.2 and 1.9.4.3) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.4.5 If there are absence or myoclonic seizures, or if JME is suspected, do not offer carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

that our results alone should be used in Pharmacological treatment of newly choosing between carbamazepine and diagnosed generalised tonic–clonic phenytoin’. [78]

(GTC) seizures A Cochrane review undertook an IPD analysis incorporating a network meta-analysis (NMA). 2018 surveillance summary Data was available from 12,391 of 17,961 A Cochrane review of IPD from 595 of 1,192 eligible participants from 36 of 77 RCTs. The eligible participants from 4 of 12 RCTs analysis compared the time to withdrawal of assessed the time to withdrawal, 6 and 12 allocated treatment, remission and first seizure month remission, and first seizure with of 10 AEDs that are currently used as carbamazepine compared to phenytoin, used monotherapy in children and adults with partial as monotherapy in people with partial onset or onset seizures or generalised tonic-clonic generalised tonic-clonic seizures, with or seizures with or without other generalised without other generalised seizure types. It seizure types (absence, myoclonus). The results found no evidence for a statistically significant indicate that carbamazepine and lamotrigine difference between carbamazepine and are suitable first-line treatments for individuals phenytoin for these outcomes, but the authors with partial onset seizures and that noted that CIs were wide and they could not levetiracetam may be a suitable alternative. exclude the possibility of important There was high-quality evidence supporting differences. There was evidence of more the use of sodium valproate as the first-line serious adverse events with phenytoin than treatment for individuals with generalised carbamazepine. While there was some tonic-clonic seizures (with or without other evidence that people with generalised seizures generalised seizure types) and that lamotrigine may be less likely to withdraw early from and levetiracetam would be suitable alternative phenytoin than from carbamazepine, first-line treatments. [80] misclassification of seizure type may have A Cochrane review of IPD available for 1,151 impacted upon the results (the authors of 1,239 eligible participants from 2 of 3 RCTs identified up to 48 individuals with evaluated the effects of topiramate misclassification of seizure type, 32% of whom monotherapy versus carbamazepine had generalised epilepsy). The review authors monotherapy for epilepsy in people with ‘recommend caution when interpreting the partial-onset seizures or generalised onset results of this review, and do not recommend tonic-clonic seizures. It reported that with

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 37 of 81 carbamazepine 12-month remission is achieved advantage of clobazam over phenytoin for earlier and it is less likely to be withdrawn than retention at six months in adolescents and topiramate in individuals with partial-onset adults with neurocysticercosis’. [84] seizures. While evidence indicated no A Cochrane review of IPD available for 480 of differences between the drugs and outcomes 517 eligible participants from 3 RCTs in people with generalised-onset tonic-clonic compared oxcarbazepine and phenytoin seizures, there were only a limited number of monotherapy in participants with partial onset individuals identified with these seizures (15% seizures or generalised tonic-clonic seizures of the sample). [81] with or without other generalised seizure A Cochrane review of IPD available for 2,572 types. It reported that there was a statistically of 3,394 eligible participants from 9 of 13 significant advantage for oxcarbazepine over RCTs evaluated the effectiveness of phenytoin for time to treatment withdrawal, lamotrigine monotherapy compared to but insufficient evidence to suggest a carbamazepine monotherapy in people with difference between the drugs for other partial onset seizures or generalised onset outcomes. By seizure type, there was no tonic-clonic seizures. It reported that in significant advantage for either drug for individuals with partial onset seizures generalised epilepsy, however oxcarbazepine is Lamotrigine was significantly less likely to be significantly less likely to be withdrawn than withdrawn than carbamazepine, but phenytoin in participants with partial onset carbamazepine may be superior in terms of seizures, but it was not possible to determine seizure control. The results were less clear for whether oxcarbazepine is equivalent, superior individuals with generalised onset seizures, or inferior to phenytoin in terms of seizure who made up only 12% of cases, as the study control. [86] authors were concerned that seizure type may A Cochrane review of IPD available for 669 of have been misclassified in up to 50% of these 1,119 eligible participants from 5 of 11 RCTs participants. Methodological quality of the investigated the time to withdrawal, remission studies was generally good. [82] and first seizure of phenytoin compared to A Cochrane review of IPD available for 836 of valproate when used as monotherapy in people 1,455 eligible participants from 6 of 13 RCTs with partial onset seizures or generalised tonic- evaluated pair-wise monotherapy comparisons clonic seizures (with or without other in people with partial onset seizures or generalised seizure types). It reported no generalised onset tonic-clonic seizures. It overall difference between the drugs for these reported that carbamazepine may be more outcomes, nor any statistical interaction effective than phenobarbitone in terms of between treatment and seizure type. The seizure control and adverse events; and that authors noted that ‘misclassification of seizure for time to first seizure recurrence type may have confounded the results of this phenobarbitone was better for partial seizures review’ and that ‘no outright evidence was and carbamazepine for generalised seizures. found to support or refute current treatment The evidence was rated as low quality. [83] policies’. [87]

A Cochrane review of 2 RCTs/quasi RCTs (163 A Cochrane review of IPD available for 599 participants) evaluating the efficacy, eligible participants from 4 of 8 RCTs effectiveness, tolerability and safety of evaluated the effectiveness of phenobarbitone clobazam as monotherapy in people with new- monotherapy versus phenytoin monotherapy onset partial or generalised seizures found ‘no in children or adults with partial onset seizures advantage for clobazam over carbamazepine or generalised onset tonic-clonic seizures. It for still taking treatment (retention) at 12 found no significant differences for seizure months in drug-naive children and a slight outcomes between the 2 AEDs, but a

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 38 of 81 statistically significant clinical advantage for Impact statement phenytoin in terms of treatment withdrawal This section of the guideline should be (the authors hypothesised that phenobarbitone updated. The new evidence supports the may lead to more adverse effects). The current recommendations concerning the substantial statistical heterogeneity between effectiveness of lamotrigine, carbamazepine studies and lack of blinding in two studies may and oxcarbazepine as first-line treatments in have confounded the results. [88] children, young people and adults with newly diagnosed generalised tonic–clonic (GTC) 2014 surveillance summary seizures. No relevant evidence was identified. While evidence identified in this surveillance Topic expert feedback and additional review also supports the effectiveness of sodium valproate, safety concerns about its information use in women have suggested the need to As discussed previously (see recommendation review the recommendations for sodium 1.9.1) concerns about the safety of sodium valproate and its place in the management of valproate treatment in women have triggered the epilepsies (see section on recommendation this exceptional review and potential update of 1.9.1). the guideline. In addition, topic experts noted that recommendation 1.9.4.1 may need to be There is also evidence to support the use of revised or re-worded due to new data and clobazam, levetiracetam and topiramate changes in policy concerning the use of sodium monotherapy in treating GTC seizures. The valproate in treating girls and women with new evidence concerning the effectiveness of epilepsy. phenytoin (not currently recommended) as a treatment in children, young people and adults On-going research, Trial comparing the with GTC seizures should be considered during effectiveness and cost effectiveness of the update of this section. levetiracetam and zonisamide versus standard treatments for epilepsy: a comparison of New evidence identified that may change Standard and New Antiepileptic Drugs current recommendations. (SANAD-II trial) was also identified (see recommendation 1.9.3).

Recommendations in this section of the guideline 1.9.5 Pharmacological treatment of absence seizures First-line treatment in children, young people and adults with absence seizures 1.9.5.1 Offer ethosuximide or sodium valproate as first-line treatment to children, young people and adults with absence seizures. If there is a high risk of GTC seizures, offer sodium valproate first, unless it is unsuitable. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.5.2 Offer lamotrigine** if ethosuximide and sodium valproate are unsuitable, ineffective or not tolerated. [new 2012] Adjunctive treatment in children, young people and adults with absence seizures 1.9.5.3 If two first-line AEDs (see recommendations 1.9.5.1 and 1.9.5.2) are ineffective in children, young people and adults with absence seizures, consider a combination of two of these three AEDs as adjunctive treatment: ethosuximide, lamotrigine[15] or sodium

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 39 of 81 valproate. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.5.4 If adjunctive treatment (see recommendation 1.9.5.3) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist and consider clobazam**, clonazepam, levetiracetam**, topiramate** or zonisamide**. [new 2012] 1.9.5.5 Do not offer carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

valproate treatment in women have triggered Pharmacological treatment of absence this exceptional review and potential update of seizures the guideline. In addition, a topic expert indicated that recommendation 1.9.5.1 may 2018 surveillance summary need to be revised or re-worded due to new A Cochrane review of 8 RCTs (n=684 data and changes in policy concerning the use evaluated the effects of ethosuximide, of sodium valproate in treating girls and valproate and lamotrigine as treatments for women with epilepsy. children and adolescents with absence seizures, when compared with either placebo Impact statement or each other. The review authors suggested The new evidence supports the that ethosuximide was the best initial empirical recommendation content within 1.9.5.1- monotherapy but valproate may be more 1.9.5.3. However safety concerns about appropriate, if absence and generalised tonic- sodium valproate in women have suggested clonic seizures coexist as ethosuximide is less the need to review the recommendations for effective for management of tonic-clonic sodium valproate and its place in the seizures. [105] management of the epilepsies (see section on recommendation 1.9.1). 2014 surveillance summary

No relevant evidence was identified. New evidence identified that may change current recommendations. Topic expert feedback and additional information As discussed previously (see recommendation 1.9.1) concerns about the safety of sodium

Recommendations in this section of the guideline 1.9.6 Pharmacological treatment of myoclonic seizures First-line treatment in children, young people and adults with myoclonic seizures

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 40 of 81 1.9.6.1 Offer sodium valproate as first-line treatment to children, young people and adults with newly diagnosed myoclonic seizures, unless it is unsuitable. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.6.2 Consider levetiracetam** or topiramate** if sodium valproate is unsuitable or not tolerated. Be aware that topiramate has a less favourable side-effect profile than levetiracetam and sodium valproate. [new 2012] Adjunctive treatment in children, young people and adults with myoclonic seizures 1.9.6.3 Offer levetiracetam, sodium valproate or topiramate** as adjunctive treatment to children, young people and adults with myoclonic seizures if first-line treatments (see recommendations 1.9.6.1 and 1.9.6.2) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.6.4 If adjunctive treatment (see recommendation 1.9.6.3) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist and consider clobazam**, clonazepam, piracetam or zonisamide**. [new 2012] 1.9.6.5 Do not offer carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

recommendation 1.9.6.1 may need to be Pharmacological treatment of revised or re-worded due to new data and myoclonic seizures changes in policy concerning the use of sodium valproate in treating girls and women with 2018 surveillance summary epilepsy. No relevant evidence was identified. Impact statement 2014 surveillance summary Safety concerns about sodium valproate in No relevant evidence was identified. women have suggested the need to review the recommendations for sodium valproate and its Topic expert feedback and additional place in the management of the epilepsies (see information discussion in section on recommendation 1.9.1). Concerns about the safety of sodium valproate treatment in girls and women have triggered this exceptional review and potential update of New evidence identified that may change current recommendations. the guideline (see recommendation 1.9.1). A topic expert also indicated that

Recommendations in this section of the guideline 1.9.7 Pharmacological treatment of tonic or atonic seizures First-line treatment in children, young people and adults with tonic or atonic seizures

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 41 of 81 1.9.7.1 Offer sodium valproate as first-line treatment to children, young people and adults with tonic or atonic seizures. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] Adjunctive treatment in children, young people and adults with tonic or atonic seizures 1.9.7.2 Offer lamotrigine** as adjunctive treatment to children, young people and adults with tonic or atonic seizures if first-line treatment with sodium valproate is ineffective or not tolerated. [new 2012] 1.9.7.3 Discuss with a tertiary epilepsy specialist if adjunctive treatment (see recommendation 1.9.7.2) is ineffective or not tolerated. Other AEDs that may be considered by the tertiary epilepsy specialist are rufinamide** and topiramate**. [new 2012] 1.9.7.4 Do not offer carbamazepine, gabapentin, oxcarbazepine, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

recommendation 1.9.7.1 may need to be Pharmacological treatment of tonic or revised or re-worded due to new data and a atonic seizures change in policy concerning the use of sodium valproate in treating girls and women with 2018 surveillance summary epilepsy. No relevant evidence was identified. Impact statement 2014 surveillance summary Safety concerns about sodium valproate in No relevant evidence was identified. women have suggested the need to review the recommendations for sodium valproate and its Topic expert feedback place in the management of the epilepsies (see discussion in section on recommendation Concerns about the safety of sodium valproate 1.9.1). treatment in girls and women have triggered this exceptional review and potential update of the guideline (see recommendation 1.9.1). A New evidence identified that may change topic expert also indicated that current recommendations.

Recommendations in this section of the guideline 1.9.8 Pharmacological treatment of infantile spasms First-line treatment in infants with infantile spasms 1.9.8.1 Discuss with, or refer to, a tertiary paediatric epilepsy specialist when an infant presents with infantile spasms. [new 2012]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 42 of 81 1.9.8.2 Offer a steroid (prednisolone or tetracosactide**) or vigabatrin as first-line treatment to infants with infantile spasms that are not due to tuberous sclerosis. Carefully consider the risk–benefit ratio when using vigabatrin or steroids. [new 2012] 1.9.8.3 Offer vigabatrin as first-line treatment to infants with infantile spasms due to tuberous sclerosis. If vigabatrin is ineffective, offer a steroid (prednisolone or tetracosactide**). Carefully consider the risk–benefit ratio when using vigabatrin or steroids. [new 2012]

* At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

Pharmacological treatment of infantile Topic expert feedback spasms Topic experts identified Safety and effectiveness of hormonal treatment versus 2018 surveillance summary hormonal treatment with vigabatrin for infantile spasms (ICISS): a randomised, A Cochrane review of RCTs of add-on multicentre, open-label trial. Topic experts said sulthiame in people of any age with any type of that recommendation 1.9.8.2 should be epilepsy found only 1 study: participants with a updated in relation to evidence on the use of a new diagnosis of West syndrome (n=37) given combination of vigabatrin and prednisolone. sulthiame as an add-on therapy to pyridoxine. Sulthiame led to a cessation of seizures. The Impact statement included study had a significant risk of bias. [106] This section of the guideline should be updated. The new evidence from Cochrane A Cochrane review that included 1 randomised reviews [107, 108] supports the control cross-over trial (5 participants) recommendations for offering prednisolone or evaluating the efficacy and tolerability of tetracosactide as a treatment for infantile corticosteroids compared to placebo or other spasms, however a recent large international antiepileptic drugs in children with epilepsy study published in 2017 demonstrated that the (excluding epileptic spasms) found no new combination of vigabatrin and prednisolone evidence for the efficacy of corticosteroids in was more effective than either oral treating childhood epilepsies. [107] prednisolone or intramuscular tetracosactide A Cochrane review of 18 RCTs (916 used alone in completely suppressing infantile participants with infantile spasms) investigated spasms in the short-term (42 days). There is 12 different pharmaceutical agents. The also an indication that sulthiame, which is not strongest evidence suggesting that hormonal currently recommended in NICE guideline treatment (prednisolone or tetracosactide CG137, may lead to a cessation of seizures depot) leads to resolution of spasms faster and when used as an add-on therapy to pyridoxine in more infants than does vigabatrin (overall in patients with West syndrome, however the poor study methodology). [108] review was based on only 1 small study with a significant risk of bias [106]. 2014 surveillance summary No relevant evidence was identified. New evidence identified that may change current recommendations.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 43 of 81 Recommendations in this section of the guideline 1.9.9 Pharmacological treatment of Dravet syndrome First-line treatment in children with Dravet syndrome 1.9.9.1 Discuss with, or refer to, a tertiary paediatric epilepsy specialist when a child presents with suspected Dravet syndrome. [new 2012] 1.9.9.2 Consider sodium valproate or topiramate* as first-line treatment in children with Dravet syndrome. [new 2012] Adjunctive treatment in children, young people and adults with Dravet syndrome 1.9.9.3 Discuss with a tertiary epilepsy specialist if first-line treatments (see recommendation 1.9.9.2) in children, young people and adults with Dravet syndrome are ineffective or not tolerated, and consider clobazam* or stiripentol as adjunctive treatment. [new 2012] 1.9.9.4 Do not offer carbamazepine, gabapentin, lamotrigine, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

(CBD) reduces convulsive seizure frequency in Pharmacological treatment of Dravet Dravet syndrome: results of a multi-center, syndrome randomized, controlled trial reported that ‘CBD resulted in greater reduction in seizure 2018 surveillance summary frequency than placebo; adverse events were A Cochrane review of 2 RCTs (64 participants) more frequent with CBD, but it was generally evaluating the effectiveness and tolerability of well-tolerated’ (there is a second publication in stiripentol in children with severe myoclonic the New England Journal of Medicine which epilepsy in infants (Dravet syndrome) found reports the same data). that it is significantly better than placebo for Initial intelligence gathering identified the achieving 50% or greater reduction in seizure proposed NICE technology appraisal (TA) frequency and seizure freedom, but there were Cannabidiol for adjuvant treatment of seizures more adverse effects. Quality of the evidence associated with Dravet syndrome or Lennox- was rated as low to moderate. [109] Gastaut syndrome that will appraise the clinical and cost effectiveness of cannabidiol within its 2014 surveillance summary marketing authorisation for adjuvant treatment No relevant evidence was identified. of seizures associated with Dravet syndrome or Lennox-Gastaut syndrome. Topic expert feedback and additional As discussed previously, NICE’s concerns about information the safety of sodium valproate treatment in A topic expert noted that there should be girls and women have triggered this recommendations on the use of cannabidiol in exceptional review and potential update of the children with refractory epilepsy, ‘as parents guideline (see recommendation 1.9.1). will often obtain this independently and there is much confusion surrounding its use, legality, Impact statement safety and efficacy’. Topic experts highlighted New evidence supports the current evidence concerning the use of cannabidiol in recommendation concerning use of stiripentol infants with Dravet syndrome: Cannabidiol in children with Dravet syndrome. There is

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 44 of 81 evidence on the use of cannabidiol for adjuvant Safety concerns about sodium valproate in girls treatment of seizures associated with Dravet and women have suggested the need to review syndrome and a planned NICE TA in this area. the recommendations for sodium valproate An editorial correction will need to be made and its place in the management of the cross-referencing to the NICE TA on epilepsies (see section on recommendation Cannabidiol (CBD) reduces convulsive seizure 1.9.1). frequency in Dravet syndrome: results of a multi-center, randomized, controlled trial when New evidence identified that may change published (if development of the TA is agreed). current recommendations.

Recommendations in this section of the guideline 1.9.10 Pharmacological treatment of Lennox–Gastaut syndrome First-line treatment in children with Lennox–Gastaut syndrome 1.9.10.1 Discuss with, or refer to, a tertiary paediatric epilepsy specialist when a child presents with suspected Lennox–Gastaut syndrome. [new 2012] 1.9.10.2 Offer sodium valproate as first-line treatment to children with Lennox–Gastaut syndrome. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] Adjunctive treatment in children, young people and adults with Lennox–Gastaut syndrome 1.9.10.3 Offer lamotrigine as adjunctive treatment to children, young people and adults with Lennox–Gastaut syndrome if first-line treatment with sodium valproate is ineffective or not tolerated. [new 2012] 1.9.10.4 Discuss with a tertiary epilepsy specialist if adjunctive treatment (see recommendation 1.9.10.3) is ineffective or not tolerated. Other AEDs that may be considered by the tertiary epilepsy specialist are rufinamide and topiramate. [new 2012] 1.9.10.5 Do not offer carbamazepine, gabapentin, oxcarbazepine, pregabalin, tiagabine or vigabatrin. [new 2012] 1.9.10.6 Only offer felbamate** in centres providing tertiary epilepsy specialist care and when treatment with all of the AEDs listed in recommendations 1.9.10.3 and 1.9.10.4 has proved ineffective or not tolerated. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

therapies used to treat Lennox-Gastaut Pharmacological treatment of Lennox– syndrome reported that rufinamide, Gastaut syndrome lamotrigine, topiramate and felbamate may be helpful as add-on therapy, and clobazam may 2018 surveillance summary be helpful for drop seizures; however the A Cochrane review of 9 RCTs (n=979) optimum treatment is unknown, and none of evaluating the effects of pharmaceutical

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 45 of 81 the studies has shown any one drug to be syndrome in paediatric and adult patients with highly efficacious. [110] treatment-resistant seizures.

2014 surveillance summary Impact statement No relevant evidence was identified. The new evidence supports the current recommendation to offer rufinamide, Topic expert feedback and additional lamotrigine, topiramate or felbamate as information adjunctive treatment for people with Lennox– No topic expert feedback was relevant to this Gastaut syndrome. Evidence for potential evidence. effectiveness of clobazam should be reviewed in an update. NICE’s concerns about the safety of sodium valproate treatment in girls and women have Safety concerns about sodium valproate in girls triggered this exceptional review and potential and women have suggested the need to review update of the guideline (see recommendation the recommendations for sodium valproate 1.9.1 for discussion of these issues). and its place in the management of the epilepsies. Initial intelligence gathering identified the in- development NICE technology appraisal (TA) An editorial correction will need to be made Cannabidiol for adjuvant treatment of seizures cross-referencing to the NICE TA on associated with Dravet syndrome or Lennox- Cannabidiol (CBD) reduces convulsive seizure Gastaut syndrome that will appraise the clinical frequency in Dravet syndrome: results of a and cost effectiveness of cannabidiol within its multi-center, randomized, controlled trial when marketing authorisation for adjuvant treatment it is published. of seizures associated with Dravet syndrome or Lennox-Gastaut syndrome. A TA is also New evidence identified that may change proposed on Cannabidiol for Lennox-Gastaut current recommendations.

Recommendations in this section of the guideline 1.9.11 Pharmacological treatment of benign epilepsy with centrotemporal spikes, Panayiotopoulos syndrome or late-onset childhood occipital epilepsy (Gastaut type) First-line treatment in children and young people with benign epilepsy with centrotemporal spikes, Panayiotopoulos syndrome or late-onset childhood occipital epilepsy (Gastaut type) 1.9.11.1 Discuss with the child or young person, and their family and/or carers, whether AED treatment for benign epilepsy with centrotemporal spikes, Panayiotopoulos syndrome or late-onset childhood occipital epilepsy (Gastaut type) is indicated. [new 2012] 1.9.11.2 Offer carbamazepine* or lamotrigine* as first-line treatment to children and young people with benign epilepsy with centrotemporal spikes, Panayiotopoulos syndrome or late- onset childhood occipital epilepsy (Gastaut type). [new 2012] 1.9.11.3 Levetiracetam is not cost effective at June 2011 unit costs**. Offer levetiracetam*, oxcarbazepine* or sodium valproate (provided the acquisition cost of levetiracetam falls to at least 50% of June 2011 value documented in the National Health Service Drug Tariff for England and Wales) if carbamazepine and lamotrigine are unsuitable or not tolerated. If the first AED tried is ineffective, offer an alternative from these five AEDs. Be aware that carbamazepine and oxcarbazepine may exacerbate or unmask continuous

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 46 of 81 spike and wave during slow sleep, which may occur in some children with benign epilepsy with centrotemporal spikes. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)†. [new 2012] 1.9.11.4 Consider adjunctive treatment if a second well-tolerated AED is ineffective (see recommendations 1.9.11.2 and 1.9.11.3). [new 2012] Adjunctive treatment in children and young people with benign epilepsy with centrotemporal spikes, Panayiotopoulos syndrome or late-onset childhood occipital epilepsy (Gastaut type) 1.9.11.5 Offer carbamazepine*, clobazam*, gabapentin*, lamotrigine*, levetiracetam*, oxcarbazepine*, sodium valproate or topiramate* as adjunctive treatment to children and young people with benign epilepsy with centrotemporal spikes, Panayiotopoulos syndrome or late-onset childhood occipital epilepsy (Gastaut type) if first-line treatments (see recommendations 1.9.11.2 and 1.9.11.3) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)†. [new 2012] 1.9.11.6 If adjunctive treatment (see recommendation 1.9.11.5) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist. Other AEDs that may be considered by the tertiary epilepsy specialist are eslicarbazepine acetate*, lacosamide*, phenobarbital, phenytoin, pregabalin*, tiagabine*, vigabatrin* and zonisamide*. Carefully consider the risk–benefit ratio when using vigabatrin because of the risk of an irreversible effect on visual fields. [new 2012]

* At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented. ** Estimated cost of a 1500 mg daily dose was £2.74 at June 2011. Cost taken from the National Health Service Drug Tariff for England and Wales. † February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients.

a low risk of bias, found that children with Pharmacological treatment of benign BECTS on sulthiame were significantly more epilepsy with centrotemporal spikes, likely to remain in seizure remission in the Panayiotopoulos syndrome or late- short-term (up to 6 months), although the onset childhood occipital epilepsy precision of the effect estimate was uncertain due to its small sample size. [111] (Gastaut type) 2014 surveillance summary 2018 surveillance summary No relevant evidence was identified. A Cochrane review of 4 RCTs (262 participants) assessing the outcome of Topic expert feedback and additional treatment with AEDs in children with Benign information Epilepsy with Centro Temporal Spikes (BECTS), reported that the studies were of low to very No topic expert feedback was relevant to this low quality and that there were no differences evidence. in seizure remission at the end of the study Concerns about the safety of sodium valproate periods between the AEDs compared treatment in women have triggered this (levetiracetam versus oxcarbazepine; clobazam exceptional review and potential update of the versus carbamazepine; carbamazepine versus guideline (see recommendation 1.9.1 for topiramate); but 1 placebo-controlled trial with discussion of these issues).

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 47 of 81 Impact statement Safety concerns about sodium valproate treatment in girls and women have suggested This section of the guideline should be the need to review the recommendations for updated. Evidence concerning the sodium valproate and its place in the effectiveness of sulthiame as a treatment in management of the epilepsies. children with Benign Epilepsy with Centro Temporal Spikes should be considered in the update. New evidence identified that may change current recommendations.

Recommendations in this section of the guideline 1.9.12 Pharmacological treatment of idiopathic generalised epilepsy (IGE) First-line treatment in children, young people and adults with IGE 1.9.12.1 Offer sodium valproate as first-line treatment to children, young people and adults with newly diagnosed IGE, particularly if there is a photoparoxysmal response on EEG. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.12.2 Offer lamotrigine** if sodium valproate is unsuitable or not tolerated. Be aware that lamotrigine can exacerbate myoclonic seizures. If JME is suspected see recommendations 1.9.13.1 and 1.9.13.2. [new 2012] 1.9.12.3 Consider topiramate** but be aware that it has a less favourable side-effect profile than sodium valproate and lamotrigine*. [new 2012] Adjunctive treatment in children, young people and adults with IGE 1.9.12.4 Offer lamotrigine**, levetiracetam**, sodium valproate or topiramate** as adjunctive treatment to children, young people and adults with IGE if first-line treatments (see recommendations 1.9.12.1, 1.9.12.2 and 1.9.12.3) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.12.5 If adjunctive treatment (see recommendation 1.9.12.4) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist and consider clobazam**, clonazepam or zonisamide**. [new 2012] 1.9.12.6 Do not offer carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 48 of 81 Concerns about the safety of sodium valproate Pharmacological treatment of treatment in women have triggered this idiopathic generalised epilepsy (IGE) exceptional review and potential update of the guideline (see recommendation 1.9.1 for 2018 surveillance summary discussion of these issues). No relevant evidence was identified. Impact statement 2014 surveillance summary This section of the guideline should be updated No relevant evidence was identified. as safety concerns about sodium valproate in women have suggested the need to review the Topic expert feedback and additional recommendations for sodium valproate and its information place in the management of the epilepsies. No topic expert feedback was relevant to this evidence. New evidence identified that may change current recommendations.

Recommendations in this section of the guideline 1.9.13 Pharmacological treatment of juvenile myoclonic epilepsy (JME) First-line treatment in children, young people and adults with JME 1.9.13.1 Offer sodium valproate as first-line treatment to children, young people and adults with newly diagnosed JME, unless it is unsuitable. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.13.2 Consider lamotrigine[15], levetiracetam[15] or topiramate[15] if sodium valproate is unsuitable or not tolerated. Be aware that topiramate has a less favourable side-effect profile than lamotrigine, levetiracetam and sodium valproate, and that lamotrigine may exacerbate myoclonic seizures. [new 2012] Adjunctive treatment in children, young people and adults with JME 1.9.13.3 Offer lamotrigine[15], levetiracetam, sodium valproate or topiramate[15] as adjunctive treatment to children, young people and adults with JME if first-line treatments (see recommendations 1.9.13.1 and 1.9.13.2) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)[14]. [new 2012] 1.9.13.4 If adjunctive treatment (see recommendation 1.9.13.3) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist and consider clobazam[14], clonazepam or zonisamide[14]. [new 2012] 1.9.13.5 Do not offer carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 49 of 81 Concerns about the safety of sodium valproate Pharmacological treatment of juvenile treatment in women have triggered this myoclonic epilepsy (JME) exceptional review and potential update of the guideline (see recommendation 1.9.1 for 2018 surveillance summary discussion of these issues). A Cochrane review of 3 RCTs (83 participants) evaluating the efficacy and tolerability of Impact statement topiramate monotherapy in the treatment of This section of the guideline should be juvenile myoclonic epilepsy found that a updated. Safety concerns about sodium greater proportion of participants in the valproate in girls and women have suggested topiramate group had a 50% or more reduction the need to review the recommendations for in primarily generalised tonic-clonic seizures sodium valproate and its place in the (PGTCS) compared with the placebo group, management of the epilepsies. While the there were no significant differences between authors of the Cochrane review concluded that topiramate versus valproate in participants there was insufficient evidence to support responding with a 50% or more reduction in topiramate for the treatment of people with myoclonic seizures or in PGTCS or seizure- JME, ‘based on the current limited available free; adverse events associated with data, topiramate seems to be better tolerated topiramate were ranked as moderate-to- than valproate’. The new evidence indicates severe, while 59% of adverse events linked to that topiramate may have a more favourable valproate were ranked as severe and systemic side-effect profile than sodium valproate, as toxicity scores were higher in the valproate such recommendation 1.9.13.2 which says that compared to topiramate group. Quality of the topiramate has a less favourable side-effect evidence was very low. [112] profile than sodium valproate, may need to be reconsidered (but evidence was considered as 2014 surveillance summary very low quality). No relevant evidence was identified. New evidence identified that may change Topic expert feedback and additional current recommendations. information No topic expert feedback was relevant to this evidence.

Recommendations in this section of the guideline 1.9.14 Pharmacological treatment of epilepsy with generalised tonic–clonic (GTC) seizures only First-line treatment in children, young people and adults with epilepsy with GTC seizures only 1.9.14.1 Offer lamotrigine or sodium valproate as first-line treatment to children, young people and adults with epilepsy with GTC seizures only. If they have suspected myoclonic seizures, or are suspected of having JME, offer sodium valproate first, unless it is unsuitable. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.14.2 Consider carbamazepine and oxcarbazepine** but be aware of the risk of exacerbating myoclonic or absence seizures. [new 2012] Adjunctive treatment in children, young people and adults with epilepsy with GTC seizures only

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 50 of 81 1.9.14.3 Offer clobazam**, lamotrigine, levetiracetam, sodium valproate or topiramate as adjunctive treatment to children, young people and adults with epilepsy with GTC seizures only, if first-line treatments (see recommendations 1.9.14.1 and 1.9.14.2) are ineffective or not tolerated. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

Concerns about the safety of sodium valproate Pharmacological treatment of epilepsy treatment in women have triggered this with generalised tonic–clonic (GTC) exceptional review and potential update of the seizures only guideline (see recommendation 1.9.1 for discussion of these issues). 2018 surveillance summary Impact statement No relevant evidence was identified. This section of the guideline should be updated 2014 surveillance summary as safety concerns about sodium valproate in women have suggested the need to review the No relevant evidence was identified. recommendations for sodium valproate and its Topic expert feedback and additional place in the management of the epilepsies. information New evidence identified that may change No topic expert feedback was relevant to this current recommendations. evidence.

Recommendations in this section of the guideline 1.9.15 Pharmacological treatment of childhood absence epilepsy, juvenile absence epilepsy or other absence epilepsy syndromes First-line treatment in children, young people and adults with childhood absence epilepsy, juvenile absence epilepsy or other absence epilepsy syndromes 1.9.15.1 Offer ethosuximide or sodium valproate as first-line treatment to children, young people and adults with absence syndromes. If there is a high risk of GTC seizures, offer sodium valproate first, unless it is unsuitable. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.15.2 Offer lamotrigine** if ethosuximide and sodium valproate are unsuitable, ineffective or not tolerated. [new 2012] Adjunctive treatment in children, young people and adults with childhood absence epilepsy, juvenile absence epilepsy or other absence epilepsy syndromes 1.9.15.3 If two first-line AEDs (see recommendations 1.9.15.1 and 1.9.15.2) are ineffective in children, young people and adults with absence epilepsy syndromes, consider a combination of two of these three AEDs as adjunctive treatment: ethosuximide,

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 51 of 81 lamotrigine** or sodium valproate. Be aware of the teratogenic and developmental risks of sodium valproate (see recommendation 1.9.1.10)*. [new 2012] 1.9.15.4 If adjunctive treatment (see recommendation 1.9.15.3) is ineffective or not tolerated, discuss with, or refer to, a tertiary epilepsy specialist and consider clobazam**, clonazepam, levetiracetam**, topiramate** or zonisamide**. [new 2012] 1.9.15.5 Do not offer carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine or vigabatrin. [new 2012]

* February 2016: see also the MHRA toolkit on the risks of valproate medicines in female patients. ** At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented.

Concerns about the safety of sodium valproate Pharmacological treatment of treatment in women have triggered this childhood absence epilepsy, juvenile exceptional review and potential update of the absence epilepsy or other absence guideline (see recommendation 1.9.1 for epilepsy syndromes discussion of these issues).

2018 surveillance summary Impact statement This section of the guideline should be updated No relevant evidence was identified. as safety concerns about sodium valproate in 2014 surveillance summary girls and women have suggested the need to review the recommendations for sodium No relevant evidence was identified. valproate and its place in the management of the epilepsies. Topic expert feedback and additional information New evidence identified that may change No topic expert feedback was relevant to this current recommendations. evidence.

Recommendations in this section of the guideline 1.9.16 Other epilepsy syndromes 1.9.16.1 Refer to a tertiary paediatric epilepsy specialist all children and young people with continuous spike and wave during slow sleep, Landau–Kleffner syndrome or myoclonic- astatic epilepsy. [new 2012]

Other epilepsy syndromes 2014 surveillance summary No relevant evidence was identified. 2018 surveillance summary No relevant evidence was identified.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 52 of 81 Topic expert feedback Impact statement No topic expert feedback was relevant to this The absence of new evidence indicates no evidence. need to update this section of the guideline.

No new evidence. Recommendations are unlikely to change.

Recommendations in this section of the guideline 1.9.17 Continuation of pharmacological treatment 1.9.17.1 Maintain a high level of vigilance for treatment-emergent adverse effects (for example, bone health issues and neuropsychiatric issues*). [new 2012] 1.9.17.2 Continuing AED therapy should be planned by the specialist. It should be part of the child, young person or adult's agreed treatment plan, which should include details of how specific drug choices were made, drug dosage, possible side effects, and action to take if seizures persist. [2004] 1.9.17.3 The needs of the child, young person or adult and their family and/or carers as appropriate should be taken into account when healthcare professionals take on the responsibility of continuing prescribing. [2004] 1.9.17.4 If management is straightforward, continuing AED therapy can be prescribed in primary care if local circumstances and/or licensing allow. [2004] 1.9.17.5 The prescriber must ensure that the child, young person or adult and their family and/or carers as appropriate are fully informed about treatment including action to be taken after a missed dose or after a gastrointestinal upset. [2004] 1.9.17.6 Adherence to treatment can be optimised with the following:  educating children, young people and adults and their families and/or carers in the understanding of their condition and the rationale of treatment  reducing the stigma associated with the condition (see also section 1.2)  using simple medication regimens  positive relationships between healthcare professionals, the child, young person or adult with epilepsy and their family and/or carers. [2004] 1.9.17.7 Regular blood test monitoring in adults is not recommended as routine, and should be done only if clinically indicated. [2004] 1.9.17.8 Regular blood test monitoring in children and young people is not recommended as routine, and should be done only if clinically indicated and recommended by the specialist. [2004] 1.9.17.9 Indications for monitoring of AED blood levels are:  detection of non-adherence to the prescribed medication  suspected toxicity  adjustment of phenytoin dose  management of pharmacokinetic interactions (for example, changes in bioavailability, changes in elimination, and co-medication with interacting drugs)

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 53 of 81  specific clinical conditions, for example, status epilepticus, organ failure and certain situations in pregnancy (see recommendation 1.15.3.19). [2012] 1.9.17.10 Examples of blood tests include:  before surgery – clotting studies in those on sodium valproate**  full blood count, electrolytes, liver enzymes, vitamin D levels, and other tests of bone metabolism (for example, serum calcium and alkaline phosphatase) every 2– 5 years for adults taking enzyme-inducing drugs. [2004] 1.9.17.11 Asymptomatic minor abnormalities in test results are not necessarily an indication for changes in medication. [2004]

* Treatment with AEDs is associated with a small risk of suicidal thoughts and behaviour; available data suggest that the increased risk applies to all AEDs and may be seen as early as 1 week after starting treatment. ** Please note that 'valproate' has been changed to 'sodium valproate' to be consistent with the terminology used in this update.

Continuation of pharmacological Topic expert feedback and additional treatment information No topic expert feedback was relevant to this 2018 surveillance summary evidence.

A Cochrane review of 12 RCTs/quasi-RCTs Intelligence gathering identified in- (1,642 participants) assessing the effectiveness development NICE guideline Behaviour of interventions aimed at improving adherence change: technology-based interventions which to antiepileptic medication in adults and may cover medication e-reminders (draft scope children with epilepsy, found that education not currently available). and counselling of participants with epilepsy resulted in mixed success (moderate-quality Impact statement evidence), behavioural interventions such as This section of the guideline should be the use of intensive reminders led to more updated. There is new evidence that indicates favourable effects (moderate-quality evidence) behavioural interventions such as intensive and mixed interventions showed improved medication reminders can be good at adherence in the intervention groups improving adherence (more so than compared to the control groups (high-quality educational interventions), but these are not evidence). [113] currently recommended within rec 1.9.17.6 on adherence. 2014 surveillance summary No relevant evidence was identified. New evidence identified that may change current recommendations.

Recommendations in this section of the guideline 1.9.18 Withdrawal of pharmacological treatment

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 54 of 81 1.9.18.1 The decision to continue or withdraw medication should be taken by the child, young person or adult, their family and/or carers as appropriate, and the specialist after a full discussion of the risks and benefits of withdrawal. At the end of the discussion children, young people and adults, and their family and/or carers as appropriate, should understand their risk of seizure recurrence on and off treatment. This discussion should take into account details of the child, young person or adult's epilepsy syndrome, prognosis and lifestyle. [2004] 1.9.18.2 Withdrawal of AEDs must be managed by, or be under the guidance of, the specialist. [2004] 1.9.18.3 The risks and benefits of continuing or withdrawing AED therapy should be discussed with children, young people and adults, and their families and/or carers as appropriate, who have been seizure free for at least 2 years (see appendix H* of the full guideline). [2004] 1.9.18.4 When AED treatment is being discontinued in a child, young person or adult who has been seizure free, it should be carried out slowly (at least 2–3 months) and one drug should be withdrawn at a time. [2004] 1.9.18.5 Particular care should be taken when withdrawing benzodiazepines and barbiturates (may take up to 6 months or longer) because of the possibility of drug-related withdrawal symptoms and/or seizure recurrence. [2004] 1.9.18.6 There should be a failsafe plan agreed with children, young people and adults and their families and/or carers as appropriate, whereby if seizures recur, the last dose reduction is reversed and medical advice is sought. [2004]

* Appendix H of the full guideline provides tables for the prognosis for remission of seizures in adults.

Withdrawal of pharmacological 2014 surveillance summary treatment No relevant evidence was identified.

2018 surveillance summary Topic expert feedback A Cochrane review of 5 RCTs (924 No topic expert feedback was relevant to this participants) assessed the risk of seizure evidence. recurrence, status epilepticus and mortality after early and late AED discontinuation in Impact statement adult and paediatric epilepsy patients. Only The new evidence supports recommendation studies with children (aged under 16 years old) 1.9.18.3, that withdrawing AED therapy should were found. The studies indicated that AEDs only be considered when seizure free for at should not be discontinued in children until least 2 years; however there is only evidence after at least two seizure-free years, for this in children, there is a lack of evidence particularly if individuals have an abnormal in adults and in children with generalised EEG (especially epileptiform activity) or partial seizures. The recommendation does not seizures, or both. There was insufficient highlight any potential concerns in withdrawing evidence to establish when to withdraw AEDs AEDs for these populations. in children with generalised seizures. The included trials were classified as low or unclear New evidence is unlikely to change guideline risk of bias. [114] recommendations.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 55 of 81

1.10 Referral for complex or refractory epilepsy

Recommendations in this section of the guideline 1.10.1 All children, young people and adults with epilepsy should have access via their specialist to a tertiary service when circumstances require. [2004] 1.10.2 If seizures are not controlled and/or there is diagnostic uncertainty or treatment failure, children, young people and adults should be referred to tertiary services soon* for further assessment. Referral should be considered when one or more of the following criteria are present:  the epilepsy is not controlled with medication within 2 years  management is unsuccessful after two drugs  the child is aged under 2 years  a child, young person or adult experiences, or is at risk of, unacceptable side effects from medication  there is a unilateral structural lesion  there is psychological and/or psychiatric co-morbidity  there is diagnostic doubt as to the nature of the seizures and/or seizure syndrome. [2004] 1.10.3 In children, the diagnosis and management of epilepsy within the first few years of life may be extremely challenging. For this reason, children with suspected epilepsy should be referred to tertiary services early, because of the profound developmental, behavioural and psychological effects that may be associated with continuing seizures. [2004] 1.10.4 Behavioural or developmental regression or inability to identify the epilepsy syndrome in a child, young person or adult should result in immediate referral to tertiary services. [2004] 1.10.5 Children, young people and adults with specific syndromes such as Sturge–Weber syndrome, the hemispheric syndromes, Rasmussen's encephalitis and hypothalamic hamartoma should be referred to a tertiary epilepsy service. [2004] 1.10.6 Psychiatric co-morbidity and/or negative baseline investigations should not be a contraindication for referral to a tertiary service**. [2004] 1.10.7 The tertiary service should include a multidisciplinary team, experienced in the assessment of children, young people and adults with complex epilepsy, and have adequate access to investigations and treatment by both medical and surgical means. [2004] 1.10.8 The expertise of multidisciplinary teams involved in managing complex epilepsy should include psychology, psychiatry, social work, occupational therapy, counselling, neuroradiology, clinical nurse specialists, neurophysiology, neurology, neurosurgery and neuroanaesthesia. Teams should have MRI and video telemetry facilities available to them. [2004] 1.10.9 The neurosurgeon in the multidisciplinary team should have specialist experience of and/or training in epilepsy surgery and have access to invasive EEG recording facilities. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 56 of 81 1.10.10 Information should be provided to children, young people and adults and families and/or carers as appropriate about the reasons for considering surgery. The benefits and risks of the surgical procedure under consideration should be fully explained before informed consent is obtained. [2004]

* The Guideline Development Group considered that 'soon' meant being seen within 4 weeks. ** In this recommendation, 'centre' has been replaced with 'service' for consistency across recommendations.

evidence on the efficacy of deep brain 2018 surveillance summary stimulation (DBS) for refractory epilepsy is No relevant evidence was identified. limited in both quantity and quality. The evidence on safety shows that there are 2014 surveillance summary serious but well-known side effects. Therefore, A RCT found that surgery improved patient this procedure should only be used with special outcomes (freedom from disabling seizures) in arrangements for clinical governance, consent, patients who had failed on 2 AEDs compared and audit or research.’ to continued AED. Patients in the surgery group also had a greater increase in HRQoL Impact statement than those in the drug-treatment group up to The evidence is consistent with the 1.5yrs following surgery. [115] recommendation in NICE guideline CG137 that patients whose epilepsy is not controlled with Topic expert feedback and additional medication within 2 years or after trying 2 information drugs should be assessed for epilepsy surgery. No topic expert feedback was relevant to this Reference to IPG416 should be considered. evidence. New evidence is unlikely to change guideline Deep brain stimulation for refractory epilepsy recommendations. (January 2012) IPG416 states that ‘the

1.11 Psychological interventions

Recommendations in this section of the guideline 1.11.1 Psychological interventions (relaxation, cognitive behaviour therapy, biofeedback) may be used in conjunction with AED therapy in adults where either the person or the specialist considers seizure control to be inadequate with optimal AED therapy. This approach may be associated with an improved quality of life in some people. [2004] 1.11.2 Psychological interventions (relaxation, cognitive behaviour therapy) may be used in children and young people with drug-resistant focal epilepsy. [2004] 1.11.3 Psychological interventions may be used as adjunctive therapy. They have not been proven to affect seizure frequency and are not an alternative to pharmacological treatment. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 57 of 81 2014 surveillance summary 2018 surveillance summary No relevant evidence was identified. A Cochrane review of 24 RCTs/quasi-RCTs (2,439 participants) assessing the effects of Topic expert feedback psychological treatments for people with epilepsy on quality of life (QoL) outcomes, No topic expert feedback was relevant to this with a meta-analysis based on data from 9 evidence. studies (468 participants), found that self- management interventions significantly Impact statement improved QoL and emotional well-being, and The guideline makes broad recommendations reduced fatigue in adults and adolescents concerning interventions to improve with epilepsy. The evidence was of moderate psychological difficulties associated with quality. [116] epilepsy including recommending relaxation A Cochrane review of studies (RCTs, quasi- and cognitive behaviour therapy; and RCTs, prospective cohort controlled studies, recognises the psychological impact of epilepsy and prospective before-and-after studies) that and that it should be identified and addressed. aimed to identify and assess possible The evidence supports the current psychological and neuropsychological recommendation, and the findings of the most interventions for adults with newly diagnosed recent Cochrane review [116] may provide epilepsy found 2 RCTs, 1 on a cognitive behavioural intervention in adolescents (low additional intervention details that could be quality evidence) which reported that the reflected in the recommendation content. This intervention significantly reduced depressive section of the guideline should be updated. symptoms in people with subthreshold depressive disorder; and 1 on a specialist New evidence identified that may change nurse intervention in adults (very low quality current recommendations. evidence) which reported no significant benefit for depressive symptoms. [117]

1.12 Ketogenic diet

Recommendations in this section of the guideline 1.12.1 Refer children and young people with epilepsy whose seizures have not responded to appropriate AEDs to a tertiary paediatric epilepsy specialist for consideration of the use of a ketogenic diet. [new 2012]

greater antiepileptic efficacy for a 4:1 2018 surveillance summary ketogenic diet over lower ratios, but the 4:1 A Cochrane review of 7 RCTs (427 ketogenic diet was consistently associated with participants) evaluating the efficacy and more adverse effects; and adverse effects tolerability of ketogenic diet and ‘similar diets’ were often the reason for participants for people with epilepsy found no studies with dropping out of trials (other reasons included adults, all participants were children or lack of efficacy and non-acceptance of the adolescents. There was some evidence for diet). While the results for the use of ketogenic

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 58 of 81 diet in epilepsy are described as ‘promising’, epilepsy in children under the age of 2 years. the evidence was poor quality. [118] And a topic expert said that the evidence for the benefit of ketogenic diet in adults with 2014 surveillance summary severe refractory epilepsy should be reviewed, One RCT in India investigated the efficacy of highlighting the Cochrane review identified in the modified Atkins diet compared to a normal the search [118]. diet in children with refractory epilepsy. Mean seizure frequency was lower in children on the Impact statement modified Atkins diet compared with the control New evidence is unlikely to impact on group. However, this non blinded study guideline recommendations due to limitations included a large proportion of vegetarian in study design and lack of sufficient evidence. children, which may not be mirrored in UK children. [119] New evidence is unlikely to change guideline recommendations. Topic expert feedback Topic experts identified an on-going RCT investigating ketogenic diet in the treatment of

1.13 Vagus nerve stimulation (VNS)

Recommendations in this section of the guideline 1.13.1 Vagus nerve stimulation is indicated for use as an adjunctive therapy in reducing the frequency of seizures in adults who are refractory to antiepileptic medication but who are not suitable for resective surgery. This includes adults whose epileptic disorder is dominated by focal seizures* (with or without secondary generalisation) or generalised seizures. [2004, amended 2012] 1.13.2 Vagus nerve stimulation is indicated for use as an adjunctive therapy in reducing the frequency of seizures in children and young people who are refractory to antiepileptic medication but who are not suitable for resective surgery. This includes children and young people whose epileptic disorder is dominated by focal seizures* (with or without secondary generalisation) or generalised seizures**. [2004, amended 2012]

* In this recommendation, 'partial seizures' has been replaced with 'focal seizures' to reflect a change in terminology since the original guideline was published in 2004. ** Evidence from Vagus nerve stimulation for refractory epilepsy in children, NICE interventional procedure guidance 50 (2004).

children with drug-resistant partial seizures and 2018 surveillance summary who are not eligible for surgery or who have A Cochrane review of 5 RCTs (439 failed surgery, found that VNS for partial participants) evaluating the effects of vagus seizures is effective and well tolerated, VNS nerve stimulation (VNS) on seizures in adults or stimulation using the high stimulation paradigm

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 59 of 81 was significantly better than low stimulation in Impact statement reducing frequency of seizures. Adverse This section of the guideline should be effects associated with implantation and updated. Evidence indicates that for focal stimulation were mainly ‘hoarseness, cough, seizures VNS appears to be effective and well dyspnea, pain, paresthesia, nausea and tolerated and that VNS stimulation using a high headache, with hoarseness and dyspnea more stimulation paradigm is significantly better than likely to occur on high stimulation than low low stimulation in reducing frequency of stimulation’ (evidence on these outcomes was seizures, however there was limited limited and of moderate to low quality). [120] information available 'so important differences between high and low stimulation cannot be 2014 surveillance summary excluded'. VNS is recommended in the No relevant evidence was identified. guideline for people with refractory epilepsy and focal seizures but there is no mention of Topic expert feedback and additional low versus high stimulation, as such the information evidence in the review may impact on the No topic expert feedback was relevant to this recommendation. evidence.

On-going research on the French vagus nerve New evidence identified that may change current recommendations. stimulation (VNS) epilepsy registry to guide physicians and their patients in the use of the VNS therapy for patients with epilepsy was identified.

1.14 Prolonged or repeated seizures and convulsive status epilepticus

Recommendations in this section of the guideline 1.14.1 First-line treatment for children, young people and adults with prolonged or repeated generalised, convulsive (tonic–clonic, tonic or clonic) seizures in the community 1.14.1.1 Give immediate emergency care and treatment to children, young people and adults who have prolonged (lasting 5 minutes or more) or repeated (three or more in an hour) convulsive seizures in the community. [2012] 1.14.1.2 Only prescribe buccal midazolam or rectal diazepam* for use in the community for children, young people and adults who have had a previous episode of prolonged or serial convulsive seizures. [new 2012] 1.14.1.3 Administer buccal midazolam as first-line treatment in children, young people and adults with prolonged or repeated seizures in the community. Administer rectal diazepam* if preferred or if buccal midazolam is not available. If intravenous access is already established and resuscitation facilities are available, administer intravenous lorazepam. [new 2012] 1.14.1.4 Treatment should be administered by trained clinical personnel or, if specified by an individually agreed protocol drawn up with the specialist, by family members or carers with appropriate training. [2004] 1.14.1.5 Care must be taken to secure the child, young person or adult's airway and assess his or her respiratory and cardiac function. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 60 of 81 1.14.1.6 Depending on response to treatment, the person's situation and any personalised care plan, call an ambulance, particularly if:  the seizure is continuing 5 minutes after the emergency medication has been administered  the person has a history of frequent episodes of serial seizures or has convulsive status epilepticus, or this is the first episode requiring emergency treatment or  there are concerns or difficulties monitoring the person's airway, breathing, circulation or other vital signs. [new 2012] 1.14.2 Treatment for children, young people and adults with convulsive status epilepticus in hospital Convulsive status epilepticus 1.14.2.1 For children, young people and adults with ongoing generalised tonic–clonic seizures (convulsive status epilepticus) who are in hospital, immediately:  secure airway  give high-concentration oxygen  assess cardiac and respiratory function  check blood glucose levels and  secure intravenous access in a large vein. See also the suggested protocols in appendix F. [new 2012] 1.14.2.2 Administer intravenous lorazepam as first-line treatment in hospital in children, young people and adults with ongoing generalised tonic–clonic seizures (convulsive status epilepticus). Administer intravenous diazepam if intravenous lorazepam is unavailable, or buccal midazolam if unable to secure immediate intravenous access. Administer a maximum of two doses of the first-line treatment (including pre-hospital treatment). See also the suggested protocols in appendix F. [new 2012] 1.14.2.3 If seizures continue, administer intravenous phenobarbital or phenytoin as second-line treatment in hospital in children, young people and adults with ongoing generalised tonic– clonic seizures (convulsive status epilepticus). See also the suggested protocols in appendix F. [new 2012] Refractory convulsive status epilepticus 1.14.2.4 Follow the suggested protocols in appendix F for treating refractory convulsive status epilepticus in secondary care. [2012] 1.14.2.5 Administer intravenous midazolam*, propofol* or thiopental sodium* to treat adults with refractory convulsive status epilepticus. Adequate monitoring, including blood levels of AEDs, and critical life systems support are required. See also the suggested protocols in appendix F. [new 2012] 1.14.2.6 Administer intravenous midazolam* or thiopental sodium* to treat children and young people with refractory convulsive status epilepticus. Adequate monitoring, including blood levels of AEDs, and critical life systems support are required. See also the suggested protocols in appendix F. [2012] 1.14.2.7 As the treatment pathway progresses, the expertise of an anaesthetist/intensivist should be sought. [2004] 1.14.2.8 If either the whole protocol or intensive care is required the tertiary service should be consulted. [2004] 1.14.2.9 Regular AEDs should be continued at optimal doses and the reasons for status epilepticus should be investigated. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 61 of 81 1.14.2.10 An individual treatment pathway should be formulated for children, young people and adults who have recurrent convulsive status epilepticus. [2004] 1.14.3 Non-convulsive status epilepticus 1.14.3.1 Non-convulsive status epilepticus is uncommon and management is less urgent. A suggested guideline can be found in appendix F. [2004]

* At the time of publication (January 2012), this drug did not have UK marketing authorisation for this indication and/or population (see appendix E for details). Informed consent should be obtained and documented in line with normal standards in emergency care.

identified 1 RCT (24 participants) which was 2018 surveillance summary terminated early due to recruitment problems. A Cochrane review of 18 RCTs (2,755 It was not possible to determine the efficacy of participants) assessed the relative propofol and thiopental sodium compared to effectiveness and safety of anticonvulsants in each other in the treatment of RSE. [122] status epilepticus reported that Intravenous (IV) lorazepam is better than IV diazepam or IV 2014 surveillance summary phenytoin alone for risk of non-cessation of A cohort study conducted in France using seizures. The risk of continuation of status French protocols indicated that treating epilepticus requiring a different drug or general patients with generalised convulsive status anaesthesia was lower with IV lorazepam than epilepticus according to a guideline-based IV diazepam, but both IV lorazepam and protocol improves outcomes. [123] diazepam were better than placebo for these outcomes. Although IV lorazepam was better A systematic review which meta-analysed 3 than IV phenytoin for reducing risk of non- small poor quality studies (based in India, Israel cessation of seizures, it was not clear whether and China) compared intravenous sodium IV valproate had any benefit over IV phenytoin. valproate with either intravenous phenytoin or For pre hospital management, midazolam IM IV diazepam in hospitalised patients with appeared more effective than lorazepam IV for generalised convulsive status epilepticus was control of seizures, frequency of hospitalisation identified. No significant difference was and ICU admissions. However the Cochrane observed between IV sodium valproate and IV authors noted ‘it was unclear whether the risk phenytoin for seizure freedom at 24 hours or of recurrence of seizures differed between seizure cessation within 30 minutes of drug. treatments’. Because of the low numbers of Patients on IV sodium valproate were less studies and participants in each comparison, likely to experience adverse effects than those differences in adverse effects between on intravenous phenytoin. [124] anticonvulsants are unclear. The quality of Topic expert feedback included studies was acceptable but risk of bias could not be determined because of Topic experts highlighted the importance of 2 incomplete and selective reporting of data in on-going trials investigating treatment for the individual studies. [121] children, young people and adults with convulsive status epilepticus in hospital: A A Cochrane review evaluating the efficacy, pragmatic randomised controlled trial of adverse effects, and short- and long-term intravenous levetiracetam versus intravenous outcomes of refractory status epilepticus (RSE) phenytoin in terminating acute, prolonged treated with one of the two anaesthetic tonic clonic seizures including convulsive status agents: thiopental sodium or propofol only

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 62 of 81 epilepticus in children, the 'EcLiPSE' Study: frequency of hospitalisation and ICU Emergency treatment with Levetiracetam or admissions, however the authors reported that Phenytoin in Status Epilepticus and Convulsive ‘it was unclear whether the risk of recurrence Status Epilepticus Paediatric Trial (ConSEPT): A of seizures differed between treatments.’[121] PREDICT study comparing levetiracetam Evidence concerning midazolam IM should be versus phenytoin for management of considered in an update. The evidence convulsive status epilepticus in children. concerning IV administration of lorazepam, diazepam and phenytoin is in line with current Impact statement recommendations; however there are 2 The section on first-line treatment for children, important on-going trials investigating young people and adults with prolonged or treatment for children, young people and repeated generalised, convulsive seizures in adults with convulsive status epilepticus in the community should be updated. There is hospital. Evidence concerning the use of new evidence on the relative effectiveness and anaesthetic agents is of low quality and safety of anticonvulsants in status epilepticus, unlikely to impact current recommendations. which indicates that intramuscular administration of midazolam is more effective New evidence identified that may change than lorazepam IV for cessation of seizures, current recommendations.

1.15 Women and girls with epilepsy

Recommendations in this section of the guideline 1.15.1 Information and advice for women and girls with epilepsy 1.15.1.1 In order to enable informed decisions and choice, and to reduce misunderstandings, women and girls with epilepsy and their partners, as appropriate, must be given accurate information and counselling about contraception, conception, pregnancy, caring for children and breastfeeding, and menopause. [2004] 1.15.1.2 Information about contraception, conception, pregnancy, or menopause should be given to women and girls in advance of sexual activity, pregnancy or menopause, and the information should be tailored to their individual needs. This information should also be given, as needed, to people who are closely involved with women and girls with epilepsy. These may include her family and/or carers. [2004] 1.15.1.3 All healthcare professionals who treat, care for, or support women and girls with epilepsy should be familiar with relevant information and the availability of counselling. [2004] 1.15.1.4 Discuss with women and girls of childbearing potential (including young girls who are likely to need treatment into their childbearing years), and their parents and/or carers if appropriate, the risk of AEDs causing malformations and possible neurodevelopmental impairments in an unborn child. Assess the risks and benefits of treatment with individual drugs. There are limited data on risks to the unborn child associated with newer drugs. Specifically discuss the risk of continued use of sodium valproate to the unborn child, being aware that higher doses of sodium valproate (more than 800 mg/day) and polytherapy, particularly with sodium valproate, are associated with greater risk. [new 2012]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 63 of 81 1.15.1.5 Be aware of the latest data on the risks to the unborn child associated with AED therapy when prescribing for women and girls of present and future childbearing potential. [2012] 1.15.1.6 All women and girls on AEDs should be offered 5 mg per day of folic acid before any possibility of pregnancy. [2004] 1.15.1.7 Refer to the SPC and BNF for individual drug advice on the interactions between AEDs and hormonal replacement and contraception. [new 2012] 1.15.2 Contraception 1.15.2.1 In women of childbearing potential, the possibility of interaction with oral contraceptives should be discussed and an assessment made as to the risks and benefits of treatment with individual drugs. [2004] 1.15.2.2 In girls of childbearing potential, including young girls who are likely to need treatment into their childbearing years, the possibility of interaction with oral contraceptives should be discussed with the child and/or her carer, and an assessment made as to the risks and benefits of treatment with individual drugs. [2004] 1.15.2.3 In women and girls of childbearing potential, the risks and benefits of different contraceptive methods, including hormone-releasing IUDs, should be discussed. [2004] 1.15.2.4 If a woman or girl taking enzyme-inducing AEDs chooses to take the combined oral contraceptive pill, guidance about dosage should be sought from the SPC and current edition of the BNF. [2004, amended 2012] 1.15.2.5 The progestogen*-only pill is not recommended as reliable contraception in women and girls taking enzyme-inducing AEDs. [2004, amended 2012] 1.15.2.6 The progestogen* implant is not recommended in women and girls taking enzyme- inducing AEDs. [2004, amended 2012] 1.15.2.7 The use of additional barrier methods should be discussed with women and girls taking enzyme-inducing AEDs and oral contraception or having depot injections of progestogen*. [2004, amended 2012] 1.15.2.8 If emergency contraception is required for women and girls taking enzyme-inducing AEDs, the type and dose of emergency contraception should be in line with the SPC and current edition of the BNF. [2004, amended 2012] 1.15.2.9 Discuss with women and girls who are taking lamotrigine that the simultaneous use of any oestrogen-based contraceptive can result in a significant reduction of lamotrigine levels and lead to loss of seizure control. When a woman or girl starts or stops taking these contraceptives, the dose of lamotrigine may need to be adjusted. [new 2012] 1.15.3 Pregnancy 1.15.3.1 Women and girls with epilepsy need accurate information during pregnancy, and the possibility of status epilepticus and SUDEP should be discussed with all women and girls who plan to stop AED therapy (see section 1.9.18). [2004] 1.15.3.2 All pregnant women and girls with epilepsy should be encouraged to notify their pregnancy, or allow their clinician to notify the pregnancy, to the UK Epilepsy and Pregnancy Register. [2004] 1.15.3.3 The clinician should discuss with the woman and girl the relative benefits and risks of adjusting medication to enable her to make an informed decision. Where appropriate, the woman or girl's specialist should be consulted. [2004] 1.15.3.4 Women and girls with generalised tonic–clonic seizures should be informed that the fetus may be at relatively higher risk of harm during a seizure, although the absolute risk remains very low, and the level of risk may depend on seizure frequency. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 64 of 81 1.15.3.5 Women and girls should be reassured that there is no evidence that focal[20], absence and myoclonic seizures affect the pregnancy or developing fetus adversely unless they fall and sustain an injury. [2004, amended 2012] 1.15.3.6 Women and girls should be reassured that an increase in seizure frequency is generally unlikely in pregnancy or in the first few months after birth. [2004] 1.15.3.7 Generally, women and girls may be reassured that the risk of a tonic–clonic seizure during the labour and the 24 hours after birth is low (1–4%). [2004] 1.15.3.8 Women and girls with epilepsy should be informed that although they are likely to have healthy pregnancies, their risk of complications during pregnancy and labour is higher than for women and girls without epilepsy. [2004] 1.15.3.9 Care of pregnant women and girls should be shared between the obstetrician and the specialist. [2004] 1.15.3.10 Pregnant women and girls who are taking AEDs should be offered a high-resolution ultrasound scan to screen for structural anomalies. This scan should be performed at 18– 20 weeks' gestation by an appropriately trained ultrasonographer, but earlier scanning may allow major malformations to be detected sooner. [2004] 1.15.3.11 The risk of seizures during labour is low, but it is sufficient to warrant the recommendation that delivery should take place in an obstetric unit with facilities for maternal and neonatal resuscitation and treating maternal seizures. [2004] 1.15.3.12 All children born to mothers taking enzyme-inducing AEDs should be given 1 mg of vitamin K parenterally at delivery. [2004] 1.15.3.13 Genetic counselling should be considered if one partner has epilepsy, particularly if the partner has idiopathic epilepsy and a positive family history of epilepsy. [2004] 1.15.3.14 Although there is an increased risk of seizures in children of parents with epilepsy, children, young people and adults with epilepsy should be given information that the probability that a child will be affected is generally low. However, this will depend on the family history. [2004] 1.15.3.15 Advanced planning, including the development of local protocols for care, should be implemented in obstetric units that deliver babies of women and girls with epilepsy. [2004] 1.15.3.16 Joint epilepsy and obstetric clinics may be convenient for mothers and healthcare professionals but there is insufficient evidence to recommend their routine use. [2004] 1.15.3.17 It is, however, important that there should be regular follow-up, planning of delivery, and liaison between the specialist or epilepsy team and the obstetrician or midwife. [2004] 1.15.3.18 Aim for seizure freedom before conception and during pregnancy (particularly for women and girls with generalised tonic–clonic seizures) but consider the risk of adverse effects of AEDs and use the lowest effective dose of each AED, avoiding polytherapy if possible. [new 2012] 1.15.3.19 Do not routinely monitor AED levels during pregnancy. If seizures increase or are likely to increase, monitoring AED levels (particularly levels of lamotrigine and phenytoin, which may be particularly affected in pregnancy) may be useful when making dose adjustments. [new 2012] 1.15.4 Breastfeeding 1.15.4.1 All women and girls with epilepsy should be encouraged to breastfeed, except in very rare circumstances. Breastfeeding for most women and girls taking AEDs is generally safe and should be encouraged. However, each mother needs to be supported in the choice of feeding method that bests suits her and her family. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 65 of 81 1.15.4.2 Prescribers should consult individual drug advice in the SPC and the BNF** when prescribing AEDs for women and girls who are breastfeeding. The decision regarding AED therapy and breastfeeding should be made between the woman or girl and the prescriber, and be based on the risks and benefits of breastfeeding against the potential risks of the drug affecting the child. [2004, amended 2012] 1.15.5 After the birth 1.15.5.1 Parents of new babies or young children should be informed that introducing a few simple safety precautions may significantly reduce the risk of accidents and minimise anxiety. An approaching birth can be an ideal opportunity to review and consider the best and most helpful measures to start to ensure maximum safety for both mother and baby. [2004] 1.15.5.2 Information should be given to all parents about safety precautions to be taken when caring for the baby (see appendix D† of the full guideline). [2004] 1.15.5.3 Parents should be reassured that the risk of injury to the infant caused by maternal seizure is low. [2004]

* In this recommendation, 'progesterone' has been replaced with 'progestogen' to reflect a change in terminology since the original guideline was published in 2004. ** In this recommendation, the original referral to appendix 5 of the BNF has been removed and replaced with more up-to-date source reference material because this appendix no longer exists and has therefore become obsolete since the original guideline was published in 2004. † Appendix D of the full guideline provides a checklist for the information needs of women and girls with epilepsy, and practical information for mothers with epilepsy.

levetiracetam and lamotrigine exposure carried 2018 surveillance summary the lowest risk of overall malformation; A Cochrane review of prospective cohort however, data pertaining to specific controlled studies, cohort studies set within malformations are lacking. [125] pregnancy registries and randomised An earlier Cochrane review that included 22 controlled trials (50 studies with 31 prospective cohort studies and 6 registry based contributing to meta-analysis) assessed the studies assessed the effects of prenatal effects of prenatal exposure to AEDs on the exposure to commonly prescribed AEDs on prevalence of congenital malformations. neurodevelopmental outcomes in children. Use Children exposed to valproate, carbamazepine of sodium valproate led to a reduction in IQ or phenytoin were at a higher risk of that was ‘sufficient to affect education and malformation than children born to women occupational outcomes in later life’. There may without epilepsy and women with untreated be a dose-relationship, with higher doses of epilepsy; children exposed to phenobarbital or sodium valproate (800 to 1000 mg daily or topiramate were at a higher risk of above) associated with a poorer cognitive malformation than children born to women outcome in the child. However, it was also without epilepsy. There was no increased risk reported that for some women sodium for major malformation with lamotrigine. valproate is the most effective drug at Gabapentin, levetiracetam, oxcarbazepine, controlling seizures. There was insufficient primidone and zonisamide were not associated data about the effects of the newer AEDs with an increased risk, but there were fewer carbamazepine, lamotrigine and phenytoin. data for these medications. When AEDs were [126] compared, children exposed to valproate had the greatest risk of malformation, while

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 66 of 81 2014 surveillance summary Initial intelligence gathering identified a publication from NIHR Links between Two observational studies on the risks of antipsychotics in pregnancy and harmful sodium valproate on foetal development [127, outcomes for baby may be influenced by 128], a cohort study on sodium valproate and mother’s lifestyle which concluded that ‘studies the impact of folic acid use in reducing risks to highlight the serious adverse consequences for foetal development [129], and a cohort study children exposed to valproate in pregnancy and on the effects of breastfeeding during AED reinforce the NICE guidance on avoiding therapy on the cognitive outcomes of offspring prescribing valproate in childbearing aged [130] were identified. All findings were in line women with mental disorders.’ with current recommendations. Concerns about the safety of sodium valproate Topic expert feedback and additional treatment in girls and women have triggered information this exceptional review and potential update of Topic experts highlighted new advice from the the guideline (see discussion in section on MHRA around the use of sodium valproate for recommendation 1.9.1). management of epilepsy, particularly in women Impact statement of childbearing age. New data relates to accumulating observational data from a This section of the guideline should be updated number of national and European Pregnancy and re-evaluate the risks and benefits of Registries. These Registries report the sodium valproate treatment and reflect the outcomes of pregnancies of women with warnings around treatment of girls and women. epilepsy, both treated and un-treated with all The evidence also indicates that while sodium anti-epileptic drugs, with follow-up periods valproate carries the highest risk of causing extending many years. These have shown a foetal malformation and developmental delay, number of anti-epileptic drugs, but particularly other AEDs are associated with similar risks sodium valproate, may be associated with and this evidence should also be considered foetal teratogenesis and developmental and during the update. cognitive impairments in children born to women taking these drugs at the time of New evidence identified that may change conception and during pregnancy. current recommendations.

1.16 Children, young people and adults with learning disabilities (see also sections 1.15 and 1.17)

Recommendations in this section of the guideline 1.16.1 Diagnosis (see also section 1.5) 1.16.1.1 It can be difficult to diagnose epilepsy in children, young people and adults with learning disabilities, and so care should be taken to obtain a full clinical history. Confusion may arise between stereotypic or other behaviours and seizure activity. [2004] 1.16.1.2 It is important to have an eye witness account supplemented by corroborative evidence (for example, a video account), where possible. [2004] 1.16.1.3 Clear, unbiased reporting is essential. Witnesses may need education to describe their observations accurately. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 67 of 81 1.16.2 Investigations (see also section 1.6) 1.16.2.1 Those with learning disabilities may require particular care and attention to tolerate investigations. [2004] 1.16.2.2 Facilities should be available for imaging under anaesthesia, if necessary. [2004] 1.16.2.3 In the child or young person presenting with epilepsy and learning disability, investigations directed at determining an underlying cause should be undertaken. [2004] 1.16.3 Management (see also section 1.8) 1.16.3.1 Enable children, young people and adults who have learning disabilities, and their family and/or carers where appropriate, to take an active part in developing a personalised care plan for treating their epilepsy while taking into account any comorbidities. [new 2012] 1.16.3.2 Ensure adequate time for consultation to achieve effective management of epilepsy in children, young people and adults with learning disabilities. [new 2012] 1.16.3.3 In making a care plan for a child, young person or adult with learning disabilities and epilepsy, particular attention should be paid to the possibility of adverse cognitive and behavioural effects of AED therapy. [2004] 1.16.3.4 The recommendations on choice of treatment and the importance of regular monitoring of effectiveness and tolerability are the same for those with learning disabilities as for the general population. [2004] 1.16.3.5 Do not discriminate against children, young people and adults with learning disabilities, and offer the same services, investigations and therapies as for the general population. [new 2012] 1.16.3.6 Every therapeutic option should be explored in children, young people and adults with epilepsy in the presence or absence of learning disabilities. [2004] 1.16.3.7 Healthcare professionals should be aware of the higher risks of mortality for children, young people and adults with learning disabilities and epilepsy and discuss these with them, their families and/or carers. [2004] 1.16.3.8 All children, young people and adults with epilepsy and learning disabilities should have a risk assessment including:  bathing and showering  preparing food  using electrical equipment  managing prolonged or serial seizures  the impact of epilepsy in social settings  SUDEP  the suitability of independent living, where the rights of the child, young person or adult are balanced with the role of the carer. [2004]

people with refractory epilepsy and intellectual 2018 surveillance summary disability and adverse events experienced by A Cochrane review of 14 RCTs (1,116 those with an intellectual disability were similar participants) evaluated the efficacy of AEDs in to those in the general population (low to people with epilepsy and intellectual moderate quality evidence). [131] disabilities. AEDs reduced seizure frequency in

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 68 of 81 A Cochrane review evaluating non- settings around lack of accessible point of pharmacological interventions for people with access for specialist services or specialist epilepsy and intellectual disabilities found nursing support. only1 RCT (n=NR; unclear risk of bias), which On-going research on Epilepsy nurse trial for compared 2 surgical procedures: callosotomy adults with intellectual disabilities was with anterior temporal lobectomy versus identified. anterior temporal lobectomy. Corpus callosotomy with anterior temporal lobectomy Impact statement was more effective than anterior temporal The evidence is in line with current lobectomy alone in improving quality of life recommendations concerning pharmacological and performance on IQ tests in people with treatment in people with epilepsy and epilepsy and intellectual disabilities but no intellectual disabilities; and evidence on non- evidence was found to support superior pharmacological treatments found no benefit in seizure control for either difference between different surgical intervention. [132] procedures on outcomes. 2014 surveillance summary No relevant evidence was identified. New evidence is unlikely to change guideline recommendations. Topic expert feedback and additional information A topic expert advised that people with a learning disability may face issues in rural

1.17 Young people with epilepsy (see also section 1.15)

Recommendations in this section of the guideline 1.17.1 The physical, psychological and social needs of young people with epilepsy should always be considered by healthcare professionals. Attention should be paid to their relationships with family and friends, and at school. [2004] 1.17.2 Healthcare professionals should adopt a consulting style that allows the young person with epilepsy to participate as a partner in the consultation. [2004] 1.17.3 Decisions about medication and lifestyle issues should draw on both the expertise of the healthcare professional and the experiences, beliefs and wishes of the young person with epilepsy as well as their family and/or carers. [2004] 1.17.4 During adolescence a named clinician should assume responsibility for the ongoing management of the young person with epilepsy and ensure smooth transition of care to adult services, and be aware of the need for continuing multi-agency support. [2004] 1.17.5 Multidisciplinary services provided jointly by adult and paediatric specialists have a key role in the care of the young person with epilepsy. This can facilitate the transition from paediatric to adult services and aid in the dissemination of information. [2004] 1.17.6 Before the transition to adult services is made, diagnosis and management should be reviewed and access to voluntary organisations, such as support groups and epilepsy charities, should be facilitated. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 69 of 81 1.17.7 The information given to young people should cover epilepsy in general and its diagnosis and treatment, the impact of seizures and adequate seizure control, treatment options including side effects and risks, and the risks of injury. Other important issues to be covered are the possible consequences of epilepsy on lifestyle and future career opportunities and decisions, driving and insurance issues, social security and welfare benefit issues, sudden death and the importance of adherence to medication regimes. Information on lifestyle issues should cover recreational drugs, alcohol, sexual activity and sleep deprivation (see section 1.3). [2004] 1.17.8 The diagnosis and management of epilepsy should be reviewed during adolescence. [2004]

2018 surveillance summary Impact statement No relevant evidence was identified. The absence of new evidence indicates that there is no need to update this section of the 2014 surveillance summary guideline. No relevant evidence was identified. No new evidence. Recommendations are Topic expert feedback unlikely to change. No topic expert feedback was relevant to this evidence.

1.18 Older people with epilepsy

Recommendations in this section of the guideline 1.18.1 Do not discriminate against older people, and offer the same services, investigations and therapies as for the general population. [new 2012] 1.18.2 Pay particular attention to pharmacokinetic and pharmacodynamic issues with polypharmacy and comorbidity in older people with epilepsy. Consider using lower doses of AEDs and, if using carbamazepine, offer controlled-release carbamazepine preparations. [new 2012]

of epilepsy in people with Alzheimer's disease 2018 surveillance summary (very low quality evidence). [133] A Cochrane review of 1 RCT (95 participants) evaluating the efficacy and tolerability of the 2014 surveillance summary treatment of epilepsy for people with No relevant evidence was identified. Alzheimer's disease did not find sufficient evidence to support levetiracetam, phenobarbital or lamotrigine for the treatment

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 70 of 81 Topic expert feedback impact of polypharmacy and comorbidity in Topic experts said there should be a focus on older people with epilepsy. treatment of epilepsy in the elderly, including atypical seizures in the very elderly and which New evidence is unlikely to change guideline AED to use in those with dementia. recommendations.

Impact statement The evidence is in line with the recommendation to consider the potential

1.19 Children, young people and adults from black and minority ethnic groups

Recommendations in this section of the guideline 1.19.1 Children, young people and adults from black and minority ethnic groups may have different cultural and communication needs and these should be considered during diagnosis and management. The need for interpretation should be considered alongside other means of ensuring that a person's needs are appropriately met. [2004] 1.19.2 An interpreter should have both cultural and medical knowledge. Interpreters from the family are generally not suitable because of issues such as confidentiality, privacy, personal dignity, and accuracy of translation. [2004] 1.19.3 Information, including information about employment rights and driving, should be available in an appropriate format or through other appropriate means for children, young people and adults who do not speak or read English. [2004]

2018 surveillance summary Impact statement No relevant evidence was identified. The absence of new evidence indicates that there is no need to update this section of the 2014 surveillance summary guideline. No relevant evidence was identified. No new evidence. Recommendations are Topic expert feedback unlikely to change. No topic expert feedback was relevant to this evidence.

1.20 Review

Recommendations in this section of the guideline 1.20.1 Children, young people and adults with epilepsy should have a regular structured review and be registered with a general medical practice. [2004]

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 71 of 81 1.20.2 Adults should have a regular structured review with their GP, but depending on the person's wishes, circumstances and epilepsy, the review may be carried out by the specialist. [2004] 1.20.3 Children and young people should have a regular structured review with a specialist. [2004] 1.20.4 For adults, the maximum interval between reviews should be 1 year but the frequency of review will be determined by the person's epilepsy and their wishes. [2004] 1.20.5 For children and young people, the maximum interval between reviews should be 1 year, but the frequency of reviews should be determined by the child or young person's epilepsy and their wishes and those of the family and/or carers. The interval between reviews should be agreed between the child or young person, their family and/or carers as appropriate, and the specialist, but is likely to be between 3 and 12 months. [2004] 1.20.6 Adults should have regular reviews. In addition, access to either secondary or tertiary care should be available to ensure appropriate diagnosis, investigation and treatment if the person or clinician view the epilepsy as inadequately controlled. [2004] 1.20.7 Adults with well-controlled epilepsy may have specific medical or lifestyle issues (for example, pregnancy or drug cessation) that may need the advice of a specialist. [2004] 1.20.8 If the structured review is to be conducted by the specialist, this may be best provided in the context of a specialist clinic. [2004] 1.20.9 Treatment should be reviewed at regular intervals to ensure that children, young people and adults with epilepsy are not maintained for long periods on treatment that is ineffective or poorly tolerated and that concordance with prescribed medication is maintained. [2004] 1.20.10 Annual review should include an enquiry about side effects and a discussion of the treatment plan to ensure concordance and adherence to medication. [2004] 1.20.11 At the review, children, young people and adults should have access to: written and visual information; counselling services; information about voluntary organisations; epilepsy specialist nurses; timely and appropriate investigations; referral to tertiary services including surgery, where appropriate. [2004]

2018 surveillance summary Impact statement No relevant evidence was identified. The absence of new evidence indicates that there is no need to update this section of the 2014 surveillance summary guideline. No relevant evidence was identified. No new evidence. Recommendations are Topic expert feedback unlikely to change. No topic expert feedback was relevant to this evidence.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 72 of 81 Research recommendations

Newly diagnosed seizures (focal and generalised) – monotherapy 4.1 How do the newer AEDs compare in efficacy to the standard AEDs in the treatment of newly diagnosed epilepsy?

 Focal seizures: carbamazepine, eslicarbazepine acetate, lacosamide, lamotrigine, levetiracetam, pregabalin and zonisamide.

 Generalised seizures: lamotrigine, levetiracetam, sodium valproate and zonisamide.

Summary of findings  New evidence relevant to the research recommendation was found (see Focal (aka partial) seizures, newly diagnosed/monotherapy and Pharmacological treatment of newly diagnosed generalised tonic–clonic (GTC) seizures). An update of the related recommendations is planned.

Surveillance decision This research recommendation will be considered again at the next surveillance point.

Epilepsy syndromes 4.2 What are the initial and add-on AEDs of choice in the treatment of the epilepsy syndromes with onset in childhood, for example, myoclonic-astatic epilepsy and Dravet syndrome?

Summary of findings  New evidence relevant to the research recommendation was found (see Pharmacological treatment of Dravet syndrome, Lennox-Gastaut syndrome and benign epilepsy with centrotemporal spikes). An update of the related recommendations is planned.

Surveillance decision This research recommendation will be considered again at the next surveillance point.

Infantile spasms 4.3 Does treatment response relate to cause in infantile spasms? Does early treatment success in seizure control and resolution of the hypsarrhythmic EEG influence the long- term developmental and cognitive outcomes more than the underlying cause of the spasms?

Summary of findings  No new evidence relevant to the research recommendation was found and no ongoing studies were identified. Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 73 of 81

© NICE 2018. All rights reserved. Subject to Notice of rights (https://www.nice.org.uk/terms- andconditions#notice-of-rights). Surveillance decision This research recommendation will be considered again at the next surveillance point.

Treatment of convulsive status epilepticus (that is, not just refractory)

4.4 What is the most effective and safest AED to treat:

 established (usually lasting longer than 30 minutes) convulsive status epilepticus

 refractory convulsive status epilepticus?

Summary of findings  New evidence relevant to the research recommendation was found. An update of the related recommendation is planned.

Surveillance decision This research recommendation will be considered again at the next surveillance point.

AEDs and pregnancy 4.5 What is the malformation rate and longer term neurodevelopmental outcome of children born to mothers who have taken AEDs during pregnancy?

Summary of findings

 New evidence relevant to the research recommendation was found. An update of the related recommendation is planned.

Surveillance decision This research recommendation will be considered again at the next surveillance point.

References

1. Fraser R, Johnson E, Lashley S, Barber J, Chaytor N, Miller J, Ciechanowski P, Temkin N, Caylor L: PACES in epilepsy: Results of a self-management randomized controlled trial. Epilepsia 2015, 56(8):1264-1274. 2. Maguire MJ, Jackson CF, Marson AG, Nevitt SJ: Treatments for the prevention of Sudden Unexpected Death in Epilepsy (SUDEP). In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 3. P R, M C, S R: Risk of sudden unexpected death in epilepsy in patients given adjunctive antiepileptic treatment for refractory seizures: a meta-analysis of placebo-controlled randomised trials. The Lancet Neurology 2011, 10(11):961-968.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 74 of 81

© NICE 2018. All rights reserved. Subject to Notice of rights (https://www.nice.org.uk/terms- andconditions#notice-of-rights). 4. Josephson CB, Sandy S, Jette N, Sajobi TT, Marshall D, Wiebe S: A systematic review of clinical decision rules for epilepsy. Epilepsy Behav 2016, 57(Pt A):69-76. 5. Keezer MR, Bouma HK, Wolfson C: The diagnostic accuracy of screening questionnaires for the identification of adults with epilepsy: a systematic review. Epilepsia 2014, 55(11):1772- 1780. 6. Asadi-Pooya AA, Rabiei AH, Tinker J, Tracy J: Review of systems questionnaire helps differentiate psychogenic nonepileptic seizures from epilepsy. J Clin Neurosci 2016, 34:105- 107. 7. Kerr WT, Janio EA, Braesch CT, Le JM, Hori JM, Patel AB, Barritt SE, Gallardo NL, Bauirjan J, Chau AM et al: Diagnostic implications of review-of-systems questionnaires to differentiate epileptic seizures from psychogenic seizures. Epilepsy Behav 2017, 69:69-74. 8. Robles L, Chiang S, Haneef Z: Review-of-systems questionnaire as a predictive tool for psychogenic nonepileptic seizures. Epilepsy Behav 2015, 45:151-154. 9. Del Bene VA, Arce Renteria M, Maiman M, Slugh M, Gazzola DM, Nadkarni SS, Barr WB: Increased odds and predictive rates of MMPI-2-RF scale elevations in patients with psychogenic non-epileptic seizures and observed sex differences. Epilepsy Behav 2017, 72:43- 50. 10. del Barrio A, Jimenez-Huete A, Toledano R, Garcia-Morales I, Gil-Nagel A: Validity of the clinical and content scales of the Multiphasic Personality Inventory Minnesota 2 for the diagnosis of psychogenic non-epileptic seizures. Neurologia 2016, 31(2):106-112. 11. Reuber M, Chen M, Jamnadas-Khoda J, Broadhurst M, Wall M, Grunewald RA, Howell SJ, Koepp M, Parry S, Sisodiya S et al: Value of patient-reported symptoms in the diagnosis of transient loss of consciousness.[Erratum appears in Neurology. 2016 Nov 15;87(20):2175; PMID: 27906663]. Neurology 2016, 87(6):625-633. 12. Kerr WT, Janio EA, Braesch CT, Le JM, Hori JM, Patel AB, Gallardo NL, Bauirjan J, D'Ambrosio SR, Chau AM et al: Identifying psychogenic seizures through comorbidities and medication history. Epilepsia 2017, 58(11):1852-1860. 13. Zou R, Wang S, Zhu L, Wu L, Lin P, Li F, Xie Z, Li X, Wang C: Calgary score and modified Calgary score in the differential diagnosis between neurally mediated syncope and epilepsy in children. Neurol Sci 2017, 38(1):143-149. 14. Spritzer SD, Pirotte BD, Agostini SD, Aniles E, Hentz JG, Noe KH, Drazkowski JF: The influence of staffing on diagnostic yield of EMU admissions: a comparison study between two institutions. Epilepsy Behav 2014, 41:264-267. 15. Sargolzaei S, Cabrerizo M, Sargolzaei A, Noei S, Eddin A, Rajaei H, Pinzon-Ardila A, Gonzalez- Arias SM, Jayakar P, Adjouadi M: A probabilistic approach for pediatric epilepsy diagnosis using brain functional connectivity networks. BMC Bioinformatics 2015, 16 Suppl 7:S9. 16. Grant AC, Abdel-Baki SG, Weedon J, Arnedo V, Chari G, Koziorynska E, Lushbough C, Maus D, McSween T, Mortati KA et al: EEG interpretation reliability and interpreter confidence: a large single-center study. Epilepsy Behav 2014, 32:102-107. 17. Fergus P, Hignett D, Hussain A, Al-Jumeily D, Abdel-Aziz K: Automatic epileptic seizure detection using scalp EEG and advanced artificial intelligence techniques. Biomed Res Int 2015, 2015:986736. 18. Goenka A, Boro A, Yozawitz E: Assessing quantitative EEG spectrograms to identify non- epileptic events. Epileptic Disord 2017, 19(3):299-306. 19. Moura LM, Shafi MM, Ng M, Pati S, Cash SS, Cole AJ, Hoch DB, Rosenthal ES, Westover MB: Spectrogram screening of adult EEGs is sensitive and efficient. Neurology 2014, 83(1):56-64. 20. Ontiveros S, Novy J, Jequier-Gygax M, Rossetti AO: Yield of Outpatient Sleep EEG for Epileptiform Alterations' Detection in Children. J Clin Neurophysiol 2016, 33(1):72-75. 21. Geut I, Weenink S, Knottnerus ILH, van Putten MJAM: Detecting interictal discharges in first seizure patients: ambulatory EEG or EEG after sleep deprivation? Seizure 2017, 51:52-54.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 75 of 81 22. Renzel R, Baumann CR, Poryazova R: EEG after sleep deprivation is a sensitive tool in the first diagnosis of idiopathic generalized but not focal epilepsy. Clin Neurophysiol 2016, 127(1):209- 213. 23. Chochoi M, Tyvaert L, Derambure P, Szurhaj W: Is long-term electroencephalogram more appropriate than standard electroencephalogram in the elderly? Clin Neurophysiol 2017, 128(1):270-274. 24. Seneviratne U, Minato E, Paul E: How reliable is ictal duration to differentiate psychogenic nonepileptic seizures from epileptic seizures? Epilepsy Behav 2017, 66:127-131. 25. Siegel J, Tatum WOt: Index-finger pointing in generalized tonic-clonic seizures. Epilepsy Behav 2016, 58:18-21. 26. Siegel J, Tatum WO: Hand postures in primary and secondary generalized tonic-clonic seizures. Neurology 2016, 87(17):1802-1805. 27. Knox A, Arya R, Horn PS, Holland K: The Diagnostic Accuracy of Video Electroencephalography Without Event Capture. Pediatr Neurol 2017, 03:03. 28. Moseley BD, Dewar S, Haneef Z, Stern JM: How long is long enough? The utility of prolonged inpatient video EEG monitoring. Epilepsy Res 2015, 109:9-12. 29. Lee CH, Lim SN, Lien F, Wu T: Duration of electroencephalographic recordings in patients with epilepsy. Seizure 2013, 22(6):438-442. 30. Theitler J, Dassa D, Gandelman-Marton R: The yield of non-elective inpatient video-EEG monitoring in adults. Neurol Sci 2017, 38(6):961-965. 31. Wu X, Su Y, Liu W, Jia X, Zhang Y, Zhang X, Wang G: Diagnostic value and prognostic evaluation of dynamic video-electroencephalogram monitoring in children with epilepsy. Experimental Ther 2017, 14(3):2527-2530. 32. Ladino LD, Voll A, Dash D, Sutherland W, Hernandez-Ronquillo L, Tellez-Zenteno JF, Moien- Afshari F: StatNet Electroencephalogram: A Fast and Reliable Option to Diagnose Nonconvulsive Status Epilepticus in Emergency Setting. Can J Neurol Sci 2016, 43(2):254-260. 33. Keezer MR, Simard-Tremblay E, Veilleux M: The Diagnostic Accuracy of Prolonged Ambulatory Versus Routine EEG. Clin EEG Neurosci 2016, 47(2):157-161. 34. Rubin MN, Jeffery OJ, Fugate JE, Britton JW, Cascino GD, Worrell GA, Hocker SE, Wijdicks EF, Rabinstein AA: Efficacy of a reduced electroencephalography electrode array for detection of seizures. Neurohospitalist 2014, 4(1):6-8. 35. van Diessen E, Otte WM, Braun KP, Stam CJ, Jansen FE: Improved diagnosis in children with partial epilepsy using a multivariable prediction model based on EEG network characteristics. PLoS ONE 2013, 8(4):e59764. 36. Tolchin B, Lee JW, Pavlova M, Dworetzky BA, Sarkis RA: Diagnostic yield of ambulatory EEGs in the elderly. Clin Neurophysiol 2017, 128(7):1350-1353. 37. Whitehead K, Sherratt M, Kandler R, Lawrence S, Pang C: Photic stimulation during electroencephalography: Efficacy and safety in an unselected cohort of patients referred to UK neurophysiology departments. Seizure 2016, 34:29-34. 38. Mazzucchi E, Vollono C, Losurdo A, Testani E, Gnoni V, Di Blasi C, Giannantoni NM, Lapenta L, Brunetti V, Della Marca G: Hyperventilation in Patients With Focal Epilepsy: Electromagnetic Tomography, Functional Connectivity and Graph Theory - A Possible Tool in Epilepsy Diagnosis? J Clin Neurophysiol 2017, 34(1):92-99. 39. Wyman AJ, Mayes BN, Hernandez-Nino J, Rozario N, Beverly SK, Asimos AW: The First-Time Seizure Emergency Department Electroencephalogram Study. Ann Emerg Med 2017, 69(2):184-191.e181. 40. Yigit O, Eray O, Mihci E, Yilmaz D, Eray B, Ozkaynak S: EEG as a part of the decision-making process in the emergency department. Eur J Emerg Med 2013, 20(6):402-407. 41. Leitinger M, Trinka E, Gardella E, Rohracher A, Kalss G, Qerama E, Hofler J, Hess A, Zimmermann G, Kuchukhidze G et al: Diagnostic accuracy of the Salzburg EEG criteria for

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 76 of 81 non-convulsive status epilepticus: a retrospective study. Lancet Neurology 2016, 15(10):1054- 1062. 42. Leitinger M, Beniczky S, Rohracher A, Gardella E, Kalss G, Qerama E, Hofler J, Hess Lindberg- Larsen A, Kuchukhidze G, Dobesberger J et al: Salzburg Consensus Criteria for Non- Convulsive Status Epilepticus--approach to clinical application. Epilepsy Behav 2015, 49:158- 163. 43. Fisch L, Lascano AM, Vernaz Hegi N, Girardin F, Kapina V, Heydrich L, Rutschmann O, Sarasin F, Vargas MI, Picard F et al: Early specialized care after a first unprovoked epileptic seizure. J Neurol 2016, 263(12):2386-2394. 44. Winston GP, Micallef C, Kendell BE, Bartlett PA, Williams EJ, Burdett JL, Duncan JS: The value of repeat neuroimaging for epilepsy at a tertiary referral centre: 16 years of experience. Epilepsy Res 2013, 105(3):349-355. 45. Paldino MJ, Yang E, Jones JY, Mahmood N, Sher A, Zhang W, Hayatghaibi S, Krishnamurthy R, Seghers V: Comparison of the diagnostic accuracy of PET/MRI to PET/CT-acquired FDG brain exams for seizure focus detection: a prospective study. Pediatr Radiol 2017, 47(11):1500-1507. 46. Tews W, Weise S, Syrbe S, Hirsch W, Viehweger A, Merkenschlager A, Bertsche A, Kiess W, Bernhard MK: Is there a predictive value of EEG and MRI after a first afebrile seizure in children? Klin Padiatr 2015, 227(2):84-88. 47. Appel S, Sharan AD, Tracy JI, Evans J, Sperling MR: A comparison of occipital and temporal lobe epilepsies. Acta Neurol Scand 2015, 132(4):284-290. 48. Albadareen R, Gronseth G, Landazuri P, He J, Hammond N, Uysal U: Postictal ammonia as a biomarker for electrographic convulsive seizures: A prospective study. Epilepsia 2016, 57(8):1221-1227. 49. Kilic TY, Yesilaras M, Atilla OD, Sever M, Aksay E: Can venous blood gas analysis be used for predicting seizure recurrence in emergency department? World J Emerg Med 2014, 5(3):187- 191. 50. Matz O, Zdebik C, Zechbauer S, Bundgens L, Litmathe J, Willmes K, Schulz JB, Dafotakis M: Lactate as a diagnostic marker in transient loss of consciousness. Seizure 2016, 40:71-75. 51. Matz O, Heckelmann J, Zechbauer S, Litmathe J, Brokmann JC, Willmes K, Schulz JB, Dafotakis M: Early postictal serum lactate concentrations are superior to serum creatine kinase concentrations in distinguishing generalized tonic-clonic seizures from syncopes. Intern 2017, 12:12. 52. Struck AF, Westover MB, Hall LT, Deck GM, Cole AJ, Rosenthal ES: Metabolic Correlates of the Ictal-Interictal Continuum: FDG-PET During Continuous EEG. Neurocrit Care 2016, 24(3):324-331. 53. Uzel M, Oray NC, Bayram B, Kume T, Girgin MC, Doylan O, Saritabak E, Yanturali S: Novel biochemical marker for differential diagnosis of seizure: ischemia-modified albumin. Am J Emerg Med 2014, 32(9):962-965. 54. Wang R, Zeng GQ, Liu X, Tong RZ, Zhou D, Hong Z: Evaluation of serum matrix metalloproteinase-3 as a biomarker for diagnosis of epilepsy. J Neurol Sci 2016, 367:291-297. 55. Wang R, Zeng GQ, Tong RZ, Zhou D, Hong Z: Serum matrix metalloproteinase-2: A potential biomarker for diagnosis of epilepsy. Epilepsy Res 2016, 122:114-119. 56. Dogan EA, Unal A, Unal A, Erdogan C: Clinical utility of serum lactate levels for differential diagnosis of generalized tonic-clonic seizures from psychogenic nonepileptic seizures and syncope. Epilepsy Behav 2017, 75:13-17. 57. An N, Zhao W, Liu Y, Yang X, Chen P: Elevated serum miR-106b and miR-146a in patients with focal and generalized epilepsy. Epilepsy Res 2016, 127:311-316. 58. Brigo F, Nardone R, Ausserer H, Storti M, Tezzon F, Manganotti P, Bongiovanni LG: The diagnostic value of urinary incontinence in the differential diagnosis of seizures. Seizure 2013, 22(2):85-90.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 77 of 81 59. Bayly J, Carino J, Petrovski S, Smit M, Fernando DA, Vinton A, Yan B, Gubbi JR, Palaniswami MS, O'Brien TJ: Time-frequency mapping of the rhythmic limb movements distinguishes convulsive epileptic from psychogenic nonepileptic seizures. Epilepsia 2013, 54(8):1402-1408. 60. Beniczky S, Conradsen I, Henning O, Fabricius M, Wolf P: Automated real-time detection of tonic-clonic seizures using a wearable EMG device. Neurology 2018, 05:05. 61. Gaggero R, Pistorio A, Pignatelli S, Rossi A, Mancardi MM, Baglietto MG, Striano P, Verrotti A: Early classification of childhood focal idiopathic epilepsies: is it possible at the first seizure? Europ J Paediatr Neurol 2014, 18(3):376-380. 62. Jackson A, Teo L, Seneviratne U: Challenges in the first seizure clinic for adult patients with epilepsy. Epileptic Disord 2016, 18(3):305-314. 63. Rizvi S, Hernandez-Ronquillo L, Moien-Afshari F, Hunter G, Waterhouse K, Dash D, Tellez- Zenteno JF: Evaluating the single seizure clinic model: Findings from a Canadian Center. J Neurol Sci 2016, 367:203-210. 64. Cunha JP, Choupina HM, Rocha AP, Fernandes JM, Achilles F, Loesch AM, Vollmar C, Hartl E, Noachtar S: NeuroKinect: A Novel Low-Cost 3Dvideo-EEG System for Epileptic Seizure Motion Quantification. PLoS ONE 2016, 11(1):e0145669. 65. Duez L, Beniczky S, Tankisi H, Hansen PO, Sidenius P, Sabers A, Fuglsang-Frederiksen A: Added diagnostic value of magnetoencephalography (MEG) in patients suspected for epilepsy, where previous, extensive EEG workup was unrevealing. Clin Neurophysiol 2016, 127(10):3301-3305. 66. Jin SH, Chung CK: Electrophysiological resting-state biomarker for diagnosing mesial temporal lobe epilepsy with hippocampal sclerosis. Epilepsy Res 2017, 129:138-145. 67. Hanrahan B, Ghearing G, Urban A, Plummer C, Pan J, Hendrickson R, Bagic A, Antony A: Diagnostic accuracy of paroxysmal spells: Clinical history versus observation. Epilepsy Behav 2017, 78:73-77. 68. Jin B, Wu H, Xu J, Yan J, Ding Y, Wang ZI, Guo Y, Wang Z, Shen C, Chen Z et al: Analyzing reliability of seizure diagnosis based on semiology. Epilepsy Behav 2014, 41:197-202. 69. Ristic AJ, Mijovic K, Bukumiric Z, Vojvodic N, Jankovic S, Bascarevic V, Dukic T, Sokic D: Differential diagnosis of a paroxysmal neurological event: Do neurologists know how to clinically recognize it? Epilepsy Behav 2017, 67:77-83. 70. RS T, Sander J, Taylor R, al. e: Predictors of health-related quality of life and costs in adults with epilepsy: a systematic review. Epilepsia 2011, 52:2168–2180. 71. CB B, Vickrey B, Caplan R, al. e: Psychiatric and medical comorbidity and quality of life outcomes in childhood-onset epilepsy. Pediatrics 2011, 128(6):e1532-1543. 72. Beretta S, Carone D, Zanchi C, Bianchi E, Pirovano M, Trentini C, Padovano G, Colombo M, Cereda D, Scanziani S et al: Long-term applicability of the new ILAE definition of epilepsy. Results from the PRO-LONG study. Epilepsia 2017, 58(9):1518-1523. 73. Bradley PM, Lindsay B, Fleeman N: Care delivery and self management strategies for adults with epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 74. Leone MA, Giussani G, Nevitt SJ, Marson AG, Beghi E: Immediate antiepileptic drug treatment, versus placebo, deferred, or no treatment for first unprovoked seizure. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 75. Powell G, Saunders M, Rigby A, Marson AG: Immediate-release versus controlled-release carbamazepine in the treatment of epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 76. Brigo F, Igwe SC, Del FA: Melatonin as add-on treatment for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 77. Milburn-McNulty P, Powell G, Sills GJ, Marson AG: Sulthiame monotherapy for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 78 of 81 78. Nevitt SJ, Marson AG, Weston J, Tudur SC: Carbamazepine versus phenytoin monotherapy for epilepsy: an individual participant data review. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 79. Xiao Y, Gan L, Wang J, Luo M, Luo H: Vigabatrin versus carbamazepine monotherapy for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 80. Nevitt SJ, Sudell M, Weston J, Tudur SC, Marson AG: Antiepileptic drug monotherapy for epilepsy: a network meta-analysis of individual participant data. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 81. Nevitt SJ, Sudell M, Tudur SC, Marson AG: Topiramate versus carbamazepine monotherapy for epilepsy: an individual participant data review. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 82. Nevitt SJ, Tudur SC, Weston J, Marson AG: Lamotrigine versus carbamazepine monotherapy for epilepsy: an individual participant data review. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 83. Nevitt SJ, Marson AG, Weston J, Tudur SC: Carbamazepine versus phenobarbitone monotherapy for epilepsy: an individual participant data review. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 84. Arya R, Anand V, Garg SK, Michael BD: Clobazam monotherapy for partial-onset or generalized-onset seizures. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014. 85. Zhou Q, Zheng J, Yu L, Jia X: Pregabalin monotherapy for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2012. 86. Nolan SJ, Muller M, Tudur SC, Marson AG: Oxcarbazepine versus phenytoin monotherapy for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 87. Nevitt SJ, Marson AG, Weston J, Tudur SC: Phenytoin versus valproate monotherapy for partial onset seizures and generalised onset tonic-clonic seizures: an individual participant data review. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 88. Nolan SJ, Tudur SC, Pulman J, Marson AG: Phenobarbitone versus phenytoin monotherapy for partial onset seizures and generalised onset tonic-clonic seizures. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 89. Xiao Y, Luo M, Wang J, Luo H: Losigamone add-on therapy for partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 90. Hemming K, Maguire MJ, Hutton JL, Marson AG: Vigabatrin for refractory partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 91. Weston J, Shukralla A, McKay AJ, Marson AG: Lacosamide add-on therapy for partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 92. Chang X-C, Yuan H, Wang Y, Xu H-Q, Hong W-K, Zheng R-Y: Eslicarbazepine acetate add-on for drug-resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 93. Shi LL, Dong J, Ni H, Geng J, Wu T: Felbamate as an add-on therapy for refractory partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 94. Al-Bachari S, Pulman J, Hutton JL, Marson AG: Gabapentin add-on for drug-resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 95. Ramaratnam S, Panebianco M, Marson AG: Lamotrigine add-on for drug-resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 96. Mbizvo GK, Dixon P, Hutton JL, Marson AG: Levetiracetam add-on for drug-resistant focal epilepsy: an updated Cochrane Review. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2012. 97. Pulman J, Hemming K, Marson AG: Pregabalin add-on for drug-resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 79 of 81 98. Brigo F, Storti M, Igwe SC: Stiripentol for focal refractory epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 99. Pulman J, Hutton JL, Marson AG: Tiagabine add-on for drug-resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014. 100. Pulman J, Jette N, Dykeman J, Hemming K, Hutton JL, Marson AG: Topiramate add-on for drug-resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014. 101. Carmichael K, Pulman J, Lakhan SE, Parikh P, Marson AG: Zonisamide add-on for drug- resistant partial epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 102. JA F, GL K, V B, D S, H Y, A L, D K, MA. R: Adjunctive perampanel for refractory partial-onset seizures: randomized phase III study 304. Neurology 2012, 79(6):589–596 103. JA F, GL K, BJ S, D S, H Y, D K, A. L: Evaluation of adjunctive perampanel in patients with refractory partial-onset seizures: Results of randomized global phase III study 305. Epilepsia 2013, 54(1):117-125. 104. GL K, JM S, V V, M E, Z H, J F, H Y, D S, HB E, J Z et al: Randomized phase III study 306: adjunctive perampanel for refractory partial-onset seizures. Neurology 2012, 78(18):1408- 1415. 105. Brigo F, Igwe SC: Ethosuximide, sodium valproate or lamotrigine for absence seizures in children and adolescents. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 106. Milburn-McNulty P, Powell G, Sills GJ, Marson AG: Sulthiame add-on therapy for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 107. Mehta V, Ferrie CD, Cross JH, Vadlamani G: Corticosteroids including ACTH for childhood epilepsy other than epileptic spasms. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 108. Hancock EC, Osborne JP, Edwards SW: Treatment of infantile spasms. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 109. Brigo F, Igwe SC, Bragazzi NL: Antiepileptic drugs for the treatment of infants with severe myoclonic epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 110. Hancock EC, Cross JH: Treatment of Lennox-Gastaut syndrome. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2013. 111. Tan HJ, Singh J, Gupta R, de GC: Comparison of antiepileptic drugs, no treatment, or placebo for children with benign epilepsy with centro temporal spikes. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014. 112. Liu J, Wang L-N, Wang Y-P: Topiramate monotherapy for juvenile myoclonic epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 113. Al-aqeel S, Gershuni O, Al-sabhan J, Hiligsmann M: Strategies for improving adherence to antiepileptic drug treatment in people with epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 114. Strozzi I, Nolan SJ, Sperling MR, Wingerchuk DM, Sirven J: Early versus late antiepileptic drug withdrawal for people with epilepsy in remission. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 115. Engel JJ, McDermott M, Wiebe S, Langfitt J: Early surgical therapy for drug-resistant temporal lobe epilepsy: a randomized trial. JAMA 2012, 307(9):922-930. 116. Michaelis R, Tang V, Wagner JL, Modi AC, LaFrance JWC, Goldstein LH, Lundgren T, Reuber M: Psychological treatments for people with epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017.

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 80 of 81 117. Jackson CF, Makin SM, Baker GA: Neuropsychological and psychological interventions for people with newly diagnosed epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 118. Martin K, Jackson CF, Levy RG, Cooper PN: Ketogenic diet and other dietary treatments for epilepsy. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 119. Sharma S, Sankhyan N, Gulati S, Agarwala A: Use of the modified Atkins diet for treatment of refractory childhood epilepsy: a randomized controlled trial. Epilepsia 2013, 54(3):481–486. 120. Panebianco M, Rigby A, Weston J, Marson AG: Vagus nerve stimulation for partial seizures. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 121. Prasad M, Krishnan PR, Sequeira R, Al-Roomi K: Anticonvulsant therapy for status epilepticus. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014. 122. Prabhakar H, Kalaivani M: Propofol versus thiopental sodium for the treatment of refractory status epilepticus. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2017. 123. Aranda A, Foucart G, Ducassé J: Generalized convulsive status epilepticus management in adults: a cohort study with evaluation of professional practice. Epilepsia 2010, 51:2159–2167. 124. Brigo F, Storti M, Del Felice A, Fiaschi A, Bongiovanni L: IV Valproate in generalized convulsive status epilepticus: a systematic review. European Journal of Neurology 2012, 19(9):1180-1191. 125. Weston J, Bromley R, Jackson CF, Adab N, Clayton-Smith J, Greenhalgh J, Hounsome J, McKay AJ, Tudur SC, Marson AG: Monotherapy treatment of epilepsy in pregnancy: congenital malformation outcomes in the child. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016. 126. Bromley R, Weston J, Adab N, Greenhalgh J, Sanniti A, McKay AJ, Tudur SC, Marson AG: Treatment for epilepsy in pregnancy: neurodevelopmental outcomes in the child. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2014. 127. Campbell E, Devenney E, Morrow J, al. e: Recurrence risk of congenital malformations in infants exposed to antiepileptic drugs in utero. Epilepsia 2013, 54(1):165–171 128. Vajda F, O'Brien T, Lander C, al. e: Teratogenesis in repeated pregnancies in antiepileptic drug-treated women. Epilepsia 2013, 54(1): 181–186. 129. Meador K, Baker G, Browning N, al. e: Fetal antiepileptic drug exposure and cognitive outcomes at age 6 years (NEAD study): a prospective observational study. Lancet Neurology 2013, 12(3):244–252. 130. Meador K, Baker G, Browning N, al. e: Effects of breastfeeding in children of women taking antiepileptic drugs. Neurology 2010, 75(22):1954–1960. 131. Jackson CF, Makin SM, Marson AG, Kerr M: Pharmacological interventions for epilepsy in people with intellectual disabilities. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 132. Jackson CF, Makin SM, Marson AG, Kerr M: Non-pharmacological interventions for people with epilepsy and intellectual disabilities. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2015. 133. Liu J, Wang L-N, Wu L-Y, Wang Y-P: Treatment of epilepsy for people with Alzheimer's disease. In: Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd; 2016.

© NICE 2018. All rights reserved. Subject to Notice of rights (https://www.nice.org.uk/terms- andconditions#notice-of-rights).

Consultation document for 2018 surveillance of Epilepsies: diagnosis and management (2012) 81 of 81